Genomic instability

Every day your cells take thousands of hits to their DNA, and a repair crew patches most of them — but the crew slows and grows sloppy with age, so damage accumulates and increasingly drives cells to malfunction, turn cancerous, or shut down. The practical message is undramatic but solid: the same habits that protect the rest of your biology — exercise above the bare-minimum guideline, an anti-inflammatory diet, real sleep — are also the best-supported ways to look after your genome, though the evidence is mostly observational, while the molecular "repair-boosting" drugs in the headlines remain experimental.

Genomic instability is the first of the twelve hallmarks of aging, and in many ways the most upstream: it is the slow corruption of the cell's master copy of its own instructions. DNA is under constant assault from outside the body — ultraviolet light, radiation, chemical pollutants — and from inside it, where the ordinary business of metabolism throws off reactive oxygen species (unstable, oxygen-containing molecules that damage whatever they touch) and other corrosive by-products. Cells carry an elaborate set of repair systems to fix these lesions, but the systems themselves become less accurate with age. The result is a rising burden of mutations and structural damage that erodes how well cells work and pushes them toward senescence (a worn-out but un-dying state) or cancer.[1]

How confident should you be

Descriptive — Strong. Intervention — Weak.

That split is the whole story of this hallmark. That DNA damage accumulates with age and that repair capacity declines are well established, and the accelerated-aging diseases show that badly broken genome maintenance can speed aging up. Whether ordinary, everyday mutation build-up sets the pace of normal aging is less settled. People who inherit a faulty DNA-copying enzyme (POLE or POLD1) collect mutations in their normal tissues at up to about seven times the usual rate, yet apart from a higher cancer risk they show no obvious premature aging, and many live into late old age. Their larger-scale genome damage (copy-number changes, rearrangements, telomere loss) is similar to other people's, though, so the case for those kinds of damage remains open.[2] What is not established is that anything you can take or do measurably slows genomic aging in a healthy person. No human trial has shown that reducing DNA damage slows aging. The lifestyle levers below are supported by observational data on biological-age markers, and the few trials that measured DNA damage directly mostly found nothing. The molecular interventions are mostly cell-culture and mouse work plus a few small trials in patients. The one approved drug, for a rare genetic progeria, does not generalise to healthy adults.

Why the body lets its genome decay

Strong for the pattern; the evolutionary explanation is theory rather than a measurement.

It can seem strange that evolution would tolerate a genome that falls apart. The classic explanation goes back to the nineteenth-century biologist August Weismann, who drew a line between the germline — the egg and sperm cells that carry genetic information to the next generation — and the soma, the rest of the body. The germline has to carry its DNA forward indefinitely, so it maintains it far more carefully: germline mutation rates are at least ten times lower than those of body cells, and germ cells with compromised genomes are made to self-destruct rather than passed on. Somatic cells only have to last long enough to get an organism through its reproductive years; after that, the evolutionary pressure to maintain a flawless genome falls away.[3]

This is not just a tidy story. A 2022 study sequenced the genomes of 208 intestinal crypts (tiny pockets of gut lining, each descended from a single stem cell) from 16 mammal species. It found that the yearly somatic mutation rate falls as species lifespan rises: short-lived species mutate fast, long-lived ones slowly. As a result, although lifespan varied about 30-fold, the mutation load at the end of life differed only about 3-fold between species. That is what you would expect if genome maintenance is a dial evolution sets according to how long a body needs to last. The authors caution that it is a correlation, though: other forms of damage, such as telomere shortening and protein turnover, show similar links with lifespan, and non-causal explanations have not been ruled out.[4] The damage itself comes in many forms: single chemical letters of the code being altered or oxidised, the two strands of the double helix getting welded together so they can't separate, and outright breaks across both strands. When these lesions are left unrepaired, they stall the machinery that copies DNA and reads it into proteins, which trips internal alarm systems that can force the cell into senescence, into programmed death, or — if the alarms fail — toward becoming a tumour.[5]

The repair brake: why adult cells under-invest in fixing DNA

Weak / preclinical — worms, cell culture and mice only.

One of the more counter-intuitive findings of the last decade is that adult tissue doesn't just lose repair capacity passively — it actively holds it down. A conserved protein assembly, sometimes called the DREAM complex, works as a master switch — the finding was published in Nature Structural & Molecular Biology in 2023 — that keeps a large set of repair genes switched off in resting cells.[6]

The brake's assembly depends on an enzyme, a kinase called DYRK1A, so it is in principle druggable. In worms, mutations that disable the brake made somatic tissue resistant to ultraviolet light, chemical DNA damage and DNA breaks. In resting human cells, the brake sits on 67 DNA-repair genes, and the DYRK1A blocker harmine switched 58 of them back on and made the cells markedly more resistant to ultraviolet light and a DNA-damaging chemical. In mice with a repair defect that causes accelerated aging (Ercc1-deficient), two weeks of harmine reduced DNA damage and cell death in the retina.[7] A 2026 follow-up strengthened the case that this brake matters for aging itself: across 92 mammal species, lower DREAM activity marked the longer-lived ones, and in mice lacking the complex, the brain accumulated 19.6% fewer insertions and deletions in its DNA. The human data in that study are correlations, though.[8] This is genuinely interesting biology and a plausible future target, but it is worm, cell-culture and mouse work — there is no human evidence that "releasing the repair brake" is safe or beneficial, and ramping up DNA repair indiscriminately carries obvious cancer-related risks. Treat it as a research frontier, not an intervention.

Accelerated-aging diseases: the strongest evidence that the genome gates aging

Strong — this is the natural experiment, and it is the firmest ground on the page.

The most compelling reason to take genomic instability seriously as a driver of aging — rather than a passive symptom — comes from a group of rare genetic diseases that fast-forward the aging process in humans. In several of them, the broken gene encodes a helicase — an enzyme that unwinds the DNA double helix so it can be copied and repaired. When a helicase is faulty, repair fails and damage accumulates prematurely, producing the features of early aging together with a high cancer risk. Werner syndrome, Bloom syndrome, and Rothmund-Thomson syndrome are the textbook examples, catalogued in a 2025 review of nucleases and helicases in genome maintenance.[9] Cockayne syndrome, caused by faults in a repair system that fixes DNA damage in actively used genes (one of the two genes involved also encodes a helicase), shows the other face of the problem. Patients age prematurely, lose brain cells widely and waste away, but unlike the related repair disorder xeroderma pigmentosum they are not prone to cancer.[10]

A related disease attacks the genome's container rather than its repair crew. In Hutchinson-Gilford progeria syndrome (HGPS), a mutation produces a malformed version of lamin — a structural protein that lines the inside of the nucleus and helps protect the DNA — which deforms the nucleus and destabilises the genome. Children with HGPS age dramatically fast and characteristically die of cardiovascular disease in adolescence. It is also the rare case here with an approved drug: lonafarnib blocks farnesylation, a chemical tag the defective lamin needs to do its damage, and it became the first treatment ever approved for a progeria.[11] The evidence comes from small single-group trials with no placebo or untreated comparison group. In the first, 25 children took it for at least two years, and arterial stiffness improved in some of them, which the authors called preliminary evidence.[12] The survival benefit comes from comparing treated children with matched untreated children from a separate natural-history study: after a median 2.2 years of treatment, 1 of 27 treated children had died (3.7%) against 9 of 27 untreated (33.3%), a hazard ratio of 0.12, or roughly 88% lower risk of death; with so few deaths, though, the 95% confidence interval (the range where the true effect plausibly lies) ran all the way from 0.01 to 0.93. The authors note that this observational design limits interpretation.[13] An earlier matched analysis of children treated with farnesylation inhibitors put the gain in average survival at about 1.6 years.[14] The drug slows the disease. It does not cure it. A second drug aimed at the faulty lamin, progerinin, has finished a small dose-finding trial in 10 people with HGPS, though its results have not yet been posted.[15] These diseases are the clearest natural experiment available: break the machinery that maintains the genome, and aging speeds up.

From local DNA damage to whole-body inflammation

Strong for the mechanism; Weak for anything you can do about it. The pathway is well mapped in cells and animals, and it is the accepted explanation for how a per-cell problem becomes systemic. No intervention targeting it has been tested in humans for aging; the only human trial so far is a small one in skin lupus, of a drug listed as targeting cGAS.[16]

For a long time genomic instability looked like a cell-by-cell problem: a damaged cell either fixes itself, dies, or turns cancerous, and that was the end of the story. The discovery that links it to systemic aging is that broken DNA can leak out of where it belongs and trigger an inflammatory alarm that spreads far beyond the original cell.

The key player is a cytosolic sensor abbreviated cGAS-STING — essentially a burglar alarm that watches for DNA in the wrong place. DNA normally lives inside the nucleus or inside mitochondria (the cell's power plants). The cell interprets DNA floating loose in the cytoplasm — the cell's main interior fluid — as evidence of a virus, and cGAS-STING responds by switching on a powerful antiviral and inflammatory program.[17] Two age-related failures set this alarm off without any virus present. First, when chromosomes are mis-sorted during cell division, the strays get wrapped in their own little micronuclei — small, fragile secondary nuclei whose flimsy envelopes are prone to rupturing and spilling DNA into the cytoplasm. That last step has been worked out in cultured cells: micronuclei end up with too little of a nuclear-transport protein called RCC1, keep swelling, and develop gaps in their supporting shell that let them burst.[18] Second, damaged mitochondria leak their own DNA, which the same sensor detects — one of the threads that ties mitochondrial dysfunction to inflammation. Much of the detailed mechanism comes from cancer immunology,[19] but it has now been tested directly in aging tissue: in naturally old mice, DNA released from faltering mitochondria switched on the alarm in the brain's aged immune cells, and blocking STING with an experimental drug, H-151, reduced age-related inflammation in several organs and the brain and improved tissue function. The same drug damped inflammatory signals from senescent human cells and from fat-tissue samples taken from people with obesity, but it has not been given to living people.[20]

A brief burst of this signalling is useful — it helps clear out damaged and pre-cancerous cells. The problem is chronic, low-level activation that never resolves. Researchers have started calling sustained cGAS-STING signalling an "inflammatory clock," because it drives exactly the kind of persistent, sterile, low-grade inflammation that defines chronic inflammation, or inflammaging, and feeds stem-cell exhaustion and tissue scarring. That phrase comes from a review of the aging ovary, and it is a metaphor rather than a measured clock; the authors note that much of their ovarian model is itself borrowed from other tissues and still needs direct testing.[21] The body has its own off-switch — an enzyme called TREX1 that mops up stray cytoplasmic DNA before the alarm trips. How it changes with age is unsettled: levels are lower in the blood of older people (more so in autoimmune disease) and in senescent cells, but higher in the aged liver and brain, so whether a failing off-switch drives inflammaging is still an open question.[22] This pathway is the mechanistic bridge that turns a per-cell genome problem into a whole-body one.

Blood vessels: where genomic instability becomes a clinical problem

Moderate — a well-supported chain of mechanism ending in a hard outcome, but the chain itself has not been tested end to end in people.

The vascular system is where this chain of events shows up most clearly as disease. The cells lining blood vessels are exposed both to circulating DNA-damaging stressors and to constant mechanical force from blood flow, and over time the combination pushes them into senescence.[23] Senescent cells stop dividing but don't die quietly. Instead they pour out a steady mixture of inflammatory signals and tissue-dissolving enzymes known as the senescence-associated secretory phenotype (SASP) — in effect, they poison their neighbourhood.

In a blood vessel, that secretory output degrades the vessel wall, reduces the availability of nitric oxide (the molecule vessels use to relax and widen), and stiffens the arteries.[24] Arterial stiffening in turn drives up blood pressure, closing a loop in which DNA damage, senescence, and inflammation reinforce one another. This is one of the more concrete routes from a microscopic genome problem to a hard clinical outcome, and it connects genomic instability directly to the territory covered under cholesterol and blood pressure.

A second, more direct human route runs through the blood. With age, some blood stem cells pick up mutations that let them out-compete their neighbours, a condition called clonal hematopoiesis of indeterminate potential (CHIP). A 2024 meta-analysis of 88 studies, all adjusted for age, found that people with CHIP had about 34% higher all-cause mortality (hazard ratio [HR] 1.34), about 40% more cardiovascular events of any kind (HR 1.40) and about 76% more coronary heart disease (HR 1.76). Deaths from cardiovascular disease were not clearly higher overall (HR 1.09), only in people with larger clones (HR 1.31). These are associations, and the size of the clone, which gene was mutated and the person's own characteristics all shaped the risk, but CHIP is the clearest measured case of an age-acquired genome change tracking with hard outcomes in people.[25] The site covers it in more detail under clonal hematopoiesis.

What actually helps — the lifestyle levers

Moderate, and mostly observational. Apart from the progeria drug above and a few small trials in patients described below, these are the only interventions on this page with human data behind them. A few have been tested against DNA damage measured directly, and those trials mostly came up empty. A pooled analysis of trials in older or chronically ill people, most of them testing antioxidant vitamins or folate, found no significant drop in micronuclei (a standard marker of chromosome damage) for most interventions apart from green tea, and noted that only one trial had tested physical activity at all.[26] In a six-month trial in 117 care-home residents aged 65–98, twice-weekly resistance training, with or without a protein-and-vitamin supplement, left chromosome damage in their cheek cells no different from a cognitive-training comparison group.[27] What these levers are shown to do is track with slower biological aging on composite markers.

The honest summary is modest but consistent: the habits that track with slower genomic and biological aging are the same unglamorous ones that protect the rest of the body. None of this is genome-specific magic. The likely route is that a body kept in good metabolic and inflammatory order generates less DNA damage, though that has not been shown directly in trials.

Exercise — and probably more than the official minimum. Physical activity tracks with slower aging on DNA-methylation "clocks" such as GrimAge, which estimate biological versus calendar age: in a Singapore cohort of 631 older adults (median age 70), self-reported weekly activity went with a slightly younger GrimAge reading, though only in a single snapshot; over about four years of follow-up in 114 of them, no lifestyle habit predicted change in the clocks, which the authors say may reflect the small group.[28] A 2024 UK Biobank study of about 106,000 adults whose activity was self-reported, with a second dataset of about 42,000 who wore an activity tracker (accelerometer), measured biological age a different way — from composite scores built on routine blood tests, such as phenotypic age, and from telomere length. The World Health Organization minimum of 600 MET-minutes a week (a MET, or metabolic equivalent of task, is a multiple of resting energy use; 600 is roughly 150 minutes of moderate activity) did not give the largest effect on those scores; the strongest association with slower biological aging was in the highest-activity group, at 3,000 MET-minutes a week or more — about five times the minimum. Some of these scores were themselves built from mortality data, which the authors note may overstate the links, and no trial has established an optimal dose, so the direction is firmer than the number. For deaths, the picture was different: the lowest all-cause mortality went with an anti-inflammatory diet, moderate activity (600–3,000 MET-minutes a week) and healthy sleep, compared with the least healthy pattern of a pro-inflammatory diet, low activity and poor sleep — about 31% lower where activity was self-reported (hazard ratio 0.69, 95% confidence interval 0.54–0.88) and about 51% lower where it was measured by accelerometer (0.49, 0.29–0.83). Both confidence intervals sit entirely below 1.0, so neither is likely to be chance, though the second is much less precisely pinned down.[29] This is observational data, and diet and sleep were self-reported, so the usual caveat applies — people with healthier habits differ in many ways — but it converges with everything else on the site about zone 2 training, VO₂ max, and resistance training.

An anti-inflammatory, plant-forward diet. Dietary patterns rich in vegetables, polyphenols, and unsaturated fats are associated with longer telomeres and lower systemic inflammation, whereas diets high in saturated fat track with adverse epigenetic changes.[30] Several specific nutrients support genome upkeep directly — zinc is a cofactor for enzymes involved in DNA repair, and folate is needed to make thymidine, one of DNA's building blocks, and for DNA methylation[31] — and the polyamine spermidine induces autophagy, the cellular recycling process that clears damaged components, which extends lifespan in animal models. The one proper human trial of supplemental spermidine — 100 adults aged 60 to 90 with self-reported memory decline, taking about 0.9 mg a day for 12 months — found no effect on memory or on its biomarkers compared with placebo, though the authors note the dose may have been too low, so the food-derived and supplemental cases belong apart; see geroprotectors.[32] In practice this is the Mediterranean dietary pattern, optionally paired with fasting or time-restricted eating and fermented foods.

Sleep. Both DNA repair and the brain's overnight waste-clearance (glymphatic) system are thought to run hardest during sleep, which makes sleep the most mechanistically plausible genome-maintenance lever of the three. The direct human evidence is small. Among 49 doctors, those who worked overnight on-site calls had more DNA breaks in their blood cells and lower DNA-repair gene activity than colleagues who did not, even when sampled after three nights of adequate sleep, and a night of on-call sleep loss pushed breaks up and repair-gene activity down further.[33] That was a small cross-sectional study of one profession, so it shows disrupted sleep goes with more DNA damage, not that extra sleep protects the genome. The methylation-clock data are thinner than often claimed: in the Singapore cohort above, short (under 6 hours) or long (over 9 hours) sleep and daytime sleepiness were among 15 self-reported habits tested, and only smoking, weekly physical activity and daily cognitively stimulating activities showed a clear link with clock age, though few participants reported regular short or long sleep.[34]

The pharmacological frontier

Weak — mostly preclinical, with a few small trials in patients and one approved exception. Caution on the general strategy: the machinery that repairs DNA is the same machinery that restrains cancer, so "boost repair" is not a free action.

A range of drugs aim at the downstream consequences of genomic instability, and they are worth understanding precisely so they can be kept in perspective: none is established for healthy adults, and lonafarnib is approved only for progeria.

  • Senolytics — drugs designed to selectively kill senescent cells — and senomorphics such as rapamycin and the anti-inflammatory ruxolitinib, which instead quiet the SASP, are in human trials for specific conditions but are not routine interventions. The site covers them under cellular senescence and geroprotectors.
  • cGAS and STING inhibitors aim to dial down the inflammatory clock directly; the strategy under discussion is pulsed dosing that preserves the pathway's useful clean-up role while preventing chronic inflammation.[35] For aging this is still animal work: in the old-mouse study described above, eight weeks of H-151 also improved grip strength and treadmill endurance.[36] In people, VENT-03, a drug listed as targeting cGAS, is in a small placebo-controlled trial in about 24 adults with skin lupus, not in aging.[37]
  • NAD⁺ precursors (nicotinamide mononucleotide, NMN, and nicotinamide riboside, NR) are popular supplements that raise levels of nicotinamide adenine dinucleotide (NAD⁺), a molecule that fuels several DNA-repair enzymes.[38] The mechanistic case is real, but the human evidence is thin: much of the protective data come from cell experiments — one cited study, for instance, tested cisplatin damage in a single cancer cell line[39] — and from small trials in specific patient groups, not healthy-adult longevity outcomes. Raising blood NAD⁺ is easy: in one 12-week trial in healthy people, 250 mg a day of NMN raised blood NAD⁺ with no obvious adverse effects.[40] Turning that into a health benefit is another matter. In a 58-person trial in people with long COVID, 2,000 mg a day of NR raised blood NAD⁺ about 2.6- to 3.1-fold within 5 to 10 weeks but did not significantly improve cognition, fatigue, sleep or mood compared with placebo, and about half of those taking it from the start had dropped out by 20 weeks.[41] A 2026 pooled analysis of 15 NMN trials in adults, at 250–2,000 mg a day for 14 days to 24 weeks, found no increase in side effects in the 10 trials with safety data, and no clear effect on weight, fasting blood sugar, HbA1c (a long-term blood-sugar measure) or blood fats; DNA damage was not among the outcomes it analysed, and no trial ran longer than 24 weeks.[42] The most on-topic trial is in a DNA-repair disease: in Werner syndrome, nine patients took 1,000 mg of NR a day and a placebo for 26 weeks each, in turn. NR caused no serious side effects, modestly reduced arterial stiffness and shrank skin ulcers, although one blood-flow measure in the right leg got worse. Only six patients finished and the supplement maker supplied the capsules, so this is a hopeful signal for people whose repair machinery is broken, not evidence for anyone else.[43]

What this does and doesn't tell you

What it tells you: genomic instability is a real contributor to aging, not a metaphor, though how much everyday mutation build-up sets the pace of normal aging is still open — the accelerated-aging diseases show that badly broken genome maintenance speeds aging up, and the cGAS-STING pathway explains how damage in one cell becomes inflammation across the whole body. It also tells you that the protective levers you can actually pull are the familiar ones, and that "more exercise than the official minimum" is a defensible reading of the data.

What it doesn't tell you: that any supplement or drug currently slows genomic aging in healthy humans, that boosting DNA repair is straightforwardly safe (it isn't — the same machinery restrains cancer), or that the molecular targets in the research pipeline are ready for use. The gap between the elegance of the mechanism and the modesty of the proven interventions is, as everywhere in this field, the honest centre of the topic.

Further reading

  • Schumacher B, et al. The central role of DNA damage in the ageing process. Nature 2021.[44]
  • Cagan A, et al. Somatic mutation rates scale with lifespan across mammals. Nature 2022.[45]
  • Robinson PS, et al. Increased somatic mutation burdens in normal human cells due to defective DNA polymerases. Nat Genet 2021 — the human counter-example: high mutation burden without overt premature aging.[46]
  • Bujarrabal-Dueso A, et al. Targeting DNA damage in ageing: towards supercharging DNA repair. Nat Rev Drug Discov 2025.[47]
  • Bujarrabal-Dueso A, et al. The DREAM complex functions as conserved master regulator of somatic DNA-repair capacities. Nat Struct Mol Biol 2023.[48]
  • Margariti A, et al. The converging roles of nucleases and helicases in genome maintenance and the aging process. Life (Basel) 2025.[49]
  • Siametis A, et al. From genome to geroscience: how DNA damage shapes systemic decline. Bioessays 2025.[50]
  • Gordon LB, et al. Association of Lonafarnib Treatment vs No Treatment With Mortality Rate in Patients With Hutchinson-Gilford Progeria Syndrome. JAMA 2018.[51]
  • Jones KM, et al. The intersection of TREX1, cGAS, STING and the DNA damage theory of aging. Front Aging 2025.[52]
  • Gulen MF, et al. cGAS-STING drives ageing-related inflammation and neurodegeneration. Nature 2023.[53]
  • Ma Y, et al. The inflammatory clock: how cGAS-STING ticks in the aging ovary. Front Cell Dev Biol 2026 — the source of the "inflammatory clock" metaphor used above.[54]
  • Picos A, et al. Vascular senescence and aging: mechanisms, clinical implications, and therapeutic prospects. Biogerontology 2025.[55]
  • Singh J, et al. Clonal hematopoiesis of indeterminate potential as a prognostic factor: a systematic review and meta-analysis. Blood Adv 2024.[56]
  • Liu C, et al. Optimal lifestyle patterns for delaying ageing and reducing all-cause mortality: insights from the UK Biobank. Eur Rev Aging Phys Act 2024.[57]
  • Shan J, et al. Lifestyle factors and DNA methylation-based aging clocks: cross-sectional and longitudinal associations in the Singapore diet and healthy aging cohort. J Prev Alzheimers Dis 2026.[58]
  • van Galen I, et al. The use of nutritional interventions to enhance genomic stability in mice and delay aging. Nutrients 2026.[59]
  • Frias-Toral E, et al. Nutrition and longevity — diet in centenarians. J Transl Med 2026.[60]
  • Schwarz C, et al. Effects of Spermidine Supplementation on Cognition and Biomarkers in Older Adults With Subjective Cognitive Decline: A Randomized Clinical Trial. JAMA Netw Open 2022.[61]

— § —