🧬 Aging Wiki

somatic-mutation-theory-of-aging

Properties1
aliasessomatic mutation hypothesis, somatic mutation theory, Szilard-Failla hypothesis, somatic mutation theory of aging

22 min read

Somatic Mutation Theory of Aging

TL;DR

The somatic mutation theory of aging holds that the lifelong accumulation of heritable DNA sequence changes in somatic cells — fixed mutations propagated through cell division — progressively degrades tissue function and causes aging. The theory’s signature predictions are (1) somatic mutation rate should inversely correlate with species lifespan, (2) genotypes that increase mutation rate (“mutator phenotypes”) should accelerate aging, and (3) reducing the mutation rate should extend healthspan or lifespan. Proposed independently by Giovanni Failla (1958) and Leo Szilard (1959), this is one of the oldest mechanistic theories of aging; it has been substantially updated and revived by modern single-cell and cross-species sequencing. Its current status is contested: the rate-versus-lifespan prediction is supported with remarkable quantitative precision across mammals (Cagan 2022), but whether mutations are causally limiting versus merely accumulating in parallel with aging remains unproven in humans.

The claim

The core assertion is narrower than it first appears:

Heritable sequence changes in somatic DNA (fixed point mutations, small indels, structural variants) accumulate throughout life via replication errors, endogenous mutagens, and declining repair fidelity; this accumulation of altered cellular genomes is a primary rate-limiting cause of functional aging.

“Heritable” means transmitted to daughter cells through mitosis — this is the key distinction from the broader DNA-damage theory. A DNA double-strand break that is repaired (even imperfectly) but not propagated to daughter cells is a lesion, not a mutation in the somatic-mutation-theory sense. The theory concerns changes that compound across cell generations.

The theory’s three concrete predictions (each independently testable):

  1. Rate-lifespan prediction: Longer-lived species should accumulate somatic mutations more slowly per year. This is the most precisely tested prediction.
  2. Mutator-accelerates-aging prediction: Genetic or environmental conditions that elevate somatic mutation rate should produce accelerated aging phenotypes, not just increased cancer risk.
  3. Rate-reduction-extends-lifespan prediction: Interventions that specifically reduce somatic mutation rate (rather than all DNA damage broadly) should extend healthspan or lifespan.

Relationship to the DNA-damage theory of aging

This is the most common source of confusion: the somatic mutation theory is the narrower, mutation-specific subset of the DNA-damage theory (dna-damage-theory-of-aging). They share an origin (both trace to Szilard 1959) but have diverged conceptually:

Somatic mutation theoryDNA-damage theory
What accumulatesHeritable fixed sequence changes (mutations)All persistent DNA lesions: DSBs, adducts, crosslinks, replication stress, AND mutations
MechanismAltered gene products from mutant genomes downstreamDSB-triggered chromatin remodeling, senescence, DDR signaling, transcription-blockage, AND mutations
ReversibilityIrreversible (sequence change is permanent per cell)Partially reversible (lesions can be repaired; epigenetic disruption can be reset)
Progeroid predictionRepair-deficiency syndromes should show mutation excessAny DDR impairment that elevates unrepaired lesion burden should accelerate aging
Key evidenceCross-species mutation rate scaling (Cagan 2022)SIRT6 overexpression, NMR cGAS HR variant, ICE model

Both pages should be read together; this page covers the narrower sequence-change claim. The DDT page covers the broader damage-accumulation framing and the SIRT6/NMR molecular evidence. Claims about DNA lesions, DSB repair, and progeroid syndromes are NOT duplicated here — see dna-damage-theory-of-aging for that substrate.

Relationship to the information theory of aging

The information theory (information-theory-of-aging, status: active) is the most direct modern competitor. Its central claim is that aging is driven by epigenetic information loss, not by DNA sequence changes — and it makes this case explicitly in opposition to the somatic mutation theory:

  • Information theory: Aging is reversible because the underlying genetic sequence is intact (the “digital” backup); epigenetic drift (the “analog” corruption) can be restored via OSK reprogramming.
  • Somatic mutation theory: Aging is driven by irreversible sequence changes; restoring the epigenome without correcting mutations should not produce full functional rejuvenation.

The ICE experiment (Yang 2023, Cell) is directly relevant to this contrast: targeted DSBs that induced epigenetic disruption without permanent sequence damage accelerated aging phenotypes, and OSK reprogramming reversed them. This is evidence in favor of the information theory and against the strong somatic-mutation causal claim — the aging acceleration required no mutations. However, normal aging involves both processes simultaneously; whether mutations or epigenetic drift is the rate-limiting driver under physiological conditions remains unresolved 1.

The two hypotheses are not mutually exclusive — mutations may contribute to epigenetic dysregulation (driver mutations in epigenetic regulators like DNMT3A, TET2, ASXL1 in CHIP are the clearest example), and epigenetic dysfunction may elevate transcription-associated mutagenesis. But they disagree on which process is primary and causal. The information theory explicitly argues that mutations are NOT the primary driver.

Historical development

1958: Failla — radiation damage and aging

Giovanni Failla published “The Aging Process and Cancerogenesis” in the Annals of the New York Academy of Sciences in 1958, proposing that random damage to somatic cells — particularly chromosome damage — accumulates over time and causes aging by the same mechanism that radiation accelerates aging 2. Failla drew on the parallel between radiation-accelerated aging and normal aging, suggesting that endogenous ionizing radiation and chemical damage produce a continuous background of somatic mutations.

1959: Szilard — the “hit” model

Leo Szilard, the physicist who contributed to the Manhattan Project, independently formalized the somatic mutation theory in a landmark 1959 PNAS paper, “On the Nature of the Aging Process” 3. Szilard proposed that aging results from the cumulative effect of random non-lethal “hits” to chromosomes that reduce cellular function. He developed a mathematical model predicting that the distribution of aging rates across individuals would follow a Poisson process of hit accumulation. Szilard’s framing was explicitly chromosome-inactivation-centric: each “aging hit” was postulated to render all genes carried by the affected chromosome inactive — a functional inactivation of somatic cells rather than a strictly sequence-mutation model. His model was distinguished by its quantitative precision — it made statistical predictions about inter-individual aging-rate variation, not merely a directional claim. The hits were assumed to be random and independent, with a constant probability per chromosome per unit time across the lifespan.

1963–1970s: Curtis and the chromosome aberration extension

Howard Curtis extended the theory empirically by measuring chromosome aberrations (visible cytological damage) in aged tissues [^curtis1966]. Curtis’s studies suggested that chromosome damage frequency increases with age in liver cells. This extended the theory beyond point mutations to include chromosomal-level instability. While the methods were crude by modern standards, this work represented the first systematic empirical test of the accumulation prediction. unsourced — Curtis 1966 primary source not in a local paper archive; cited from secondary reviews.

1990s–2010s: Vijg and transgenic mutation reporters

Jan Vijg’s group at the Einstein College of Medicine developed transgenic lacZ mutation-reporter mice that allowed direct quantification of somatic mutation frequencies across tissues and ages. These studies confirmed that mutation burden increases with age (approximately 2–3-fold across tissues from young to old), provided tissue-specific mutation spectra, and demonstrated that highly proliferative tissues accumulate mutations faster in absolute terms. This work grounded the theory in measured mutation frequencies rather than inferred damage. See somatic-mutation-accumulation for the complete rate data.

2015–2018: Sanger Institute’s normal-tissue positive-selection programme

Martincorena et al. demonstrated that aged normal tissues harbor not just passively accumulating neutral mutations but clones bearing canonical cancer driver mutations expanding under positive selection — in histologically normal, non-cancerous tissue 45. This reframed the theory: the relevant biological phenomenon is not just bulk mutation burden but the emergence of mutant clones with altered cell fitness in aged tissue. Whether expanded driver-mutant clones (outside the hematopoietic compartment where CHIP is defined) impair tissue function is a key open question.

2022: Cagan — the cross-species rate constraint

The most precise quantitative test of Prediction 1 to date: Cagan et al. 2022 (Nature) measured somatic mutation rates in intestinal crypts across 16 mammalian species and found that mutation rate per year inversely correlates with species maximum lifespan (log-log LME regression FVE = 0.85, P = 1×10⁻⁶) 6. Crucially, the end-of-life somatic mutation burden varied only ~3-fold across species with a ~30-fold lifespan range — consistent with a ceiling model where a common mutation burden threshold is reached near species-specific end of life. For rate/burden data and the full statistical picture, see somatic-mutation-accumulation (verified 2026-05-26).

Evidence supporting

Prediction 1: Mutation rate inversely scales with lifespan across mammals

The cross-species constraint (Cagan 2022) is the strongest available evidence for the theory’s core prediction. Short-lived species (mouse, ~2.5 yr lifespan) accumulate ~796 somatic substitutions/genome/year in intestinal crypts; humans (~84 yr lifespan) accumulate ~47/year; intermediate species span the range in near-perfect log-log linear fashion. The approximately constant end-of-life burden (~2,000–5,000 substitutions/genome) is consistent with the hypothesis that somatic mutation burden is causally limiting across mammals, although it does not prove this causation. See somatic-mutation-accumulation for full species table and statistical detail 6.

DimensionStatus
Pathway conserved in humans?yes — study is directly in human tissue and cross-species
Phenotype conserved in humans?yes — quantified directly
Replicated in humans?yes — multiple independent tissue studies; rate-lifespan scaling confirmed

needs-replication — the causal interpretation of the scaling (mutations limit lifespan vs. lifespan evolved to minimize mutation burden) has not been experimentally resolved.

Prediction 2 (partial): Mutator phenotypes accelerate progeroid features

DNA repair-deficiency progeroid syndromes — Cockayne syndrome, xeroderma pigmentosum, Werner syndrome, Bloom syndrome — produce premature aging phenotypes in humans. All involve impaired DNA repair that elevates somatic mutation burden (among other consequences). However, this support is indirect: these syndromes elevate both mutations AND unrepaired lesions simultaneously, making it impossible to attribute the accelerated aging specifically to the mutation component. The DDT page (dna-damage-theory-of-aging) covers the progeroid syndrome evidence in detail; the somatic-mutation theory can claim partial credit but cannot cleanly distinguish its prediction from the broader DDT prediction.

Mouse “mutator” models more directly test this prediction: Pol-ε and Pol-δ proofreading-null mice accumulate somatic mutations at elevated rates and develop accelerated cancer incidence rather than accelerated aging per se — primarily intestinal and thymic tumors rather than multi-system aging phenotypes 7. This is the most important complication for Prediction 2: mutator mice develop cancer, not accelerated aging. The cancer-vs.-aging divergence in mutator models is a significant empirical challenge for the theory in its strong form. See somatic-mutation-accumulation for the proofreading-null data.

DimensionStatus
Pathway conserved in humans?yes — polymerase proofreading is conserved
Phenotype conserved in humans?partial — progeroid syndromes (mixed evidence; also elevate lesions beyond mutations)
Replicated in humans?no direct human mutator-aging experiment

needs-human-replication

CHIP: driver mutations driving systemic pathology

Clonal hematopoiesis of indeterminate potential (clonal-hematopoiesis) is the most clinically concrete evidence that somatic driver mutations in normal (non-cancerous) tissue cause pathological functional changes — specifically, elevated cardiovascular disease risk, systemic inflammation, and increased all-cause mortality. CHIP involves clonal expansion of HSCs bearing driver mutations in DNMT3A, TET2, ASXL1, or other epigenetic regulators; expanded clones are hyperinflammatory and alter tissue function without being cancer. This is the clearest current proof-of-concept that somatic mutations in aged normal tissue can impair function through a non-cancer mechanism — directly supporting the theory’s core claim beyond the hematopoietic compartment extrapolation.

Population-scale normal-tissue somatic mutation data

Lawson et al. 2025 (Nature) profiled somatic mutations across 1,042 oral epithelium and 371 blood samples using ultra-sensitive targeted NanoSeq, identified 46 genes under positive selection (abstract; 49 by a secondary dNdScv analysis in the body), documented oral epithelium mutation rate ~18 SNVs/diploid genome/year, and showed that tobacco and alcohol exposures accelerate clonal dynamics in normal tissue 8. This extends the Martincorena 2015/2018 programme to a population scale, confirming pervasive driver-mutant clonal dynamics in aged normal tissue as a generalizable phenomenon across tissues and individuals.

Evidence against / key complications

Mutator mice develop cancer, not accelerated aging

This is the most cited empirical complication for the theory. When mouse somatic mutation rate is specifically elevated (via proofreading-deficient polymerases), the outcome is tissue-specific cancer, not accelerated systemic aging 7. If mutations drove aging rate-limitingly, mutator animals should look old faster. The predominance of cancer over aging phenotypes in mutator models suggests that the cellular consequences of reaching high mutation burden at young age differ qualitatively from the effects of lifetime accumulation at normal rates — rapid saturation may trigger apoptosis or cancer rather than the gradual functional decline characteristic of aging. contradictory-evidence

The ICE experiment: aging without mutations

Yang et al. 2023 (Cell) showed that inducing DSBs without permanent sequence damage — the ICE system — accelerated epigenetic aging and functional decline. OSK reprogramming reversed both. This is direct evidence that aging can be driven and reversed via epigenetic processes independent of somatic mutations. Under the strict somatic-mutation theory interpretation, OSK reprogramming without correcting the accumulated mutation load should not rejuvenate function. The ICE result is the information theory’s strongest evidence and the somatic-mutation theory’s clearest challenge. See information-theory-of-aging for the full experimental account 9.

End-of-life burden constancy: correlation not causation

The Cagan 2022 finding that end-of-life somatic mutation burden is approximately constant across species is consistent with the causal hypothesis but does not prove it. An alternative interpretation: lifespan has co-evolved with mutation rate to minimize early-mortality cancer risk — natural selection constrains the mutation rate to below a cancer-causing threshold for the species-typical lifespan. Under this alternative, the scaling reflects evolutionary optimization, not a causal burden ceiling that, when reached, causes death. Distinguishing these interpretations requires reducing mutation rate in a normally aging animal and demonstrating extended lifespan — this experiment has not been done 1.

Most somatic mutations are in non-coding / neutral DNA

The human genome is ~98% non-coding; of the ~2,000–3,000 somatic mutations in an aged human colon epithelial cell, the vast majority are in regions with minimal functional consequence. Whether the small fraction in functional coding or regulatory regions is sufficient to impair cell function at physiological burdens is not experimentally established for most cell types. In neurons, an additional complication: most somatic mutations will never be transcribed (residing in non-expressed genomic regions). See somatic-mutation-accumulation § Functional consequences in non-dividing cells.

Invertebrate DDR mutants — null aging phenotypes

In C. elegans, many DNA repair gene loss-of-function mutants have normal lifespan or even mild extension. See caenorhabditis-elegans and the DDT page for the full discussion. This is most parsimoniously explained by the post-mitotic adult soma in C. elegans (minimal replication-based mutation accumulation), but it limits the universality of the theory. contradictory-evidence

Status assessment and rationale

Status: contested.

The theory’s rate-lifespan prediction (Prediction 1) is supported with quantitative precision across mammals. The CHIP evidence demonstrates that somatic driver mutations can impair normal-tissue function through non-cancer mechanisms in aged humans. The Bujarrabal-Dueso 2025 Nature Reviews Drug Discovery review — a consensus statement from Vijg, Robbins, Schumacher, and Garinis — explicitly frames genomic DNA damage (including mutations) as “a major causal mechanism of aging” 10.

However, three important counterweights prevent a higher confidence assignment:

  1. Causation unproven in mammals beyond CHIP. The rate-lifespan correlation is compatible with causation but does not prove it. The experiment needed — reduce somatic mutation rate specifically and observe lifespan extension — has not been performed.

  2. The information-theory challenge. The ICE reprogramming evidence shows aging can be driven and reversed via epigenetic processes without sequence change. This is not reconciled with the theory’s framing that irreversible sequence damage is the primary driver.

  3. Mutator models cause cancer, not accelerated aging. The phenotypic divergence in mutator animals is the most direct empirical test of the strong causal claim; it gives ambiguous results.

The theory is best understood as partially corroborated: Prediction 1 is strongly supported; Prediction 2 is supported with major caveats; Prediction 3 has never been directly tested in a mammal. The field’s current consensus leans toward a multi-mechanism model where somatic mutations, epigenetic drift, and signaling dysregulation operate in parallel as partially interacting aging drivers — none is rate-limiting alone.

Compare dna-damage-theory-of-aging (status: active) — the DDT is assessed more favorably because it encompasses epigenetic disruption downstream of DNA lesions (not just mutations), giving it a broader evidentiary substrate. The somatic-mutation theory is the more specific, and therefore more falsifiable, claim.

Key predictions and what would update this hypothesis

ObservationDirection of update
Reducing somatic mutation rate specifically (e.g., via improved polymerase fidelity, targeted error-correcting expression) extends lifespan in normally-aged mammalsStrong positive — first direct causal test of Prediction 3
Cross-tissue somatic mutation burden shows MR-causal effect on biological age acceleration in humans (GWAS instruments now feasible via population-scale sequencing)Strong positive
Long-lived mutant mouse with specifically elevated somatic mutation rate dies younger without cancer phenotypePositive for Prediction 2
OSK reprogramming in a high-mutation-burden animal fully reverses functional aging without correcting mutationsStrong negative — would confirm information-theory primacy over somatic-mutation-theory
Cross-species somatic mutation rate scaling extends to non-intestinal tissues and invertebrate organismsPositive (generalizability)
A mammal with genuinely elevated somatic mutation rate that nonetheless lives longer than its body-mass predictsWould substantially complicate or falsify Prediction 1 if confirmed
Functional impairment of aged neurons or cardiomyocytes specifically attributable to coding-region mutations (not DSBs or epigenetic drift)Positive — would directly demonstrate mechanism in post-mitotic cells

What experiments are missing

The single most important missing experiment for resolving the causal question in mammals is: engineering increased polymerase fidelity (or reduced intrinsic mutagenesis) in an otherwise normal mouse and measuring whether lifespan or healthspan is extended. This would directly test Prediction 3 without the confounds of the current progeroid syndrome literature (which elevates both lesions and mutations simultaneously).

A secondary experiment: Mendelian randomization in humans using population-scale somatic mutation burden as the exposure (now tractable with Lawson 2025 population-scale data) to test whether genetically-predicted higher mutation burden causally shortens lifespan.

  • dna-damage-theory-of-aging (status: active) — the broader parent hypothesis encompassing DNA lesions, not just mutations; DDT includes the somatic-mutation theory as a subset. The somatic-mutation theory makes a narrower, more falsifiable subset of DDT’s predictions. The distinction matters because the ICE experiment (epigenetic disruption without sequence change → aging) is evidence against the somatic-mutation theory specifically, but is compatible with DDT.
  • information-theory-of-aging (status: active) — the primary modern competitor; argues epigenetic information loss (not sequence change) drives aging; ICE + OSK reprogramming are its strongest evidence; explicitly positions somatic mutations as NOT the primary driver. The two theories are genuinely competing at the level of causal primacy.
  • free-radical-theory-of-aging (status: superseded) — historically overlapping; ROS-induced mutations were a major proposed mechanism for how free radicals drive aging; the null results of antioxidant supplementation on lifespan extension undercut the free-radical framing but do not falsify the somatic-mutation theory (mutations also accumulate from non-ROS sources).
  • hyperfunction-theory (status: active) — orthogonal: hyperfunction argues mTOR signaling overdrive drives aging, not damage accumulation of any kind; the rapamycin ITP data are the primary empirical tension with all damage-accumulation theories.
  • genomic-instability — the canonical wiki home for the mechanistic substrate; somatic mutation accumulation is one component of genomic instability alongside chromosomal instability, telomere erosion, and mtDNA damage
  • cellular-senescence — p53 activation by mutation-associated damage can trigger senescence; CHIP driver mutations produce senescence-like SASP in aged blood
  • stem-cell-exhaustion — accumulation of somatic mutations in stem-cell compartments (HSCs, intestinal stem cells) impairs self-renewal and differentiation output over time
  • somatic-mutation-accumulation (verified 2026-05-26) — canonical home for all rate and burden data; measured rates across tissues and species; Cagan 2022 cross-species constraint; Martincorena normal-tissue driver clones; mechanisms of accumulation; detection methods. This hypothesis page aggregates and cites from that process page; do NOT duplicate rate numerics here.
  • heterocephalus-glaber (verified-partial) — naked mole-rat; exceptionally low somatic mutation rate relative to body mass; multiple convergent genomic-stability mechanisms; ~31-year lifespan; strong circumstantial evidence for Prediction 1 within rodents
  • balaena-mysticetus (verified) — bowhead whale; ~211-year lifespan; enhanced DNA repair (CIRBP-mediated NHEJ + HR); most extreme mammalian lifespan outlier; supports Prediction 3 directionally (enhanced repair → longer lifespan), though mutation rate not directly quantified in Cagan 2022 (too few individuals)
  • caenorhabditis-elegans (verified-partial) — invertebrate exception; DNA repair mutants often have normal or extended lifespan; limits universality of the theory; most parsimoniously explained by post-mitotic soma
  • mus-musculus (verified, full) — dominant mammalian model; polymerase proofreading-null mutators develop cancer not accelerated aging; lacZ reporter studies established age-related mutation accumulation; CHIP models confirm driver-mutation → functional impairment mechanism

Knowledge gaps

  • #gap/needs-replication — Causal direction of rate-lifespan correlation (Cagan 2022): correlation is robust but the experiment that would distinguish causation from evolutionary coevolution has not been done.
  • #gap/needs-human-replication — No direct human experiment tests whether reducing somatic mutation rate extends healthspan; all causal evidence is indirect (progeroid syndromes, CHIP, comparative biology).
  • #gap/contradictory-evidence — Mutator mouse models (Pol-ε/Pol-δ proofreading-null) develop cancer rather than accelerated systemic aging — the primary phenotypic prediction of the theory in its strong form is not confirmed in the most direct test.
  • #gap/contradictory-evidence — ICE + OSK reprogramming (Yang 2023): aging can be driven and reversed via epigenetic processes independent of somatic mutations — directly challenges the somatic-mutation-theory’s irreversibility claim.
  • #gap/no-mechanism — Whether the ~2,000–3,000 coding-region mutations in an aged human colon cell are sufficient to impair cell function at physiological levels (vs. being mostly neutral) has not been established experimentally.
  • #gap/unsourced — Failla 1958 (doi:10.1111/j.1749-6632.1958.tb46828.x) is closed-access (not_oa); historical framing claims remain scoped to secondary review 1. Szilard 1959 is now locally downloaded and verified.
  • #gap/unsourced — Curtis 1966 primary source not in a local paper archive; cited from secondary reviews.
  • #gap/needs-replication — Population-scale somatic mutation–aging MR (Mendelian randomization using somatic burden as exposure): feasible with Lawson 2025 data but not yet performed.

Footnotes

Footnotes

  1. doi:10.1016/j.arr.2021.101316 · Vijg J · Ageing Res Rev 2021;68:101316 · review · PMID 33711511 · 106 citations · FWCI 7.8 · 100th citation percentile · OA via PMC (PMC10018438) · download pending · “From DNA damage to mutations: All roads lead to aging” — synthesis of DDT and somatic-mutation theory from Szilard 1959 through modern multi-omic data; source for historical framing of Szilard and Curtis claims. ↩ ↩2 ↩3

  2. doi:10.1111/j.1749-6632.1958.tb46828.x · Failla G · Ann N Y Acad Sci 1958;71:1124–1135 · review / hypothesis · 215 citations · FWCI 34.1 · 100th citation percentile · not_oa (closed access) · “The Aging Process and Cancerogenesis” — proposed radiation-analogy somatic mutation theory of aging; argues endogenous ionizing radiation and chemical damage produce a continuous background of somatic mutations causing aging; one of the two foundational papers of the theory. unsourced — closed access (not_oa); full text not verifiable locally; claims here rely on secondary review 1. ↩

  3. doi:10.1073/pnas.45.1.30 · Szilard L · Proc Natl Acad Sci USA 1959;45(1):30–45 · hypothesis / mathematical model · 638 citations · FWCI 22.9 · 100th citation percentile · green OA via PMC (PMCID 222509) · local PDF available · “On the Nature of the Aging Process” — formalized the somatic mutation theory as a Poisson hit model; proposed that random non-lethal “aging hits” to chromosomes render all genes on the affected chromosome inactive; accumulation of inactivated chromosomes reduces the surviving functional fraction of somatic cells, causing aging; made quantitative statistical predictions about inter-individual aging-rate variation via Poisson distribution of hit accumulation. Affiliation: Enrico Fermi Institute of Nuclear Studies, University of Chicago. ↩

  4. martincorena-2015-somatic-mutations-normal-skin · doi:10.1126/science.aaa6806 · PMID 25999502 · n=234 microdissected skin patches, 4 donors aged 55–73 · observational WGS · model: human sun-exposed epidermis · local PDF available · 1,769 citations · FWCI 53.1 · 100th citation percentile · pervasive driver-mutant clones in normal aged skin; ~25% of cells carry NOTCH1/TP53/FAT1 driver mutations; positive selection in histologically normal tissue. ↩

  5. martincorena-2018-somatic-clones-oesophagus · doi:10.1126/science.aau3879 · PMID 30337457 · n=844 biopsies, 9 individuals aged 20–75 · observational WGS + microdissection · model: human normal oesophageal epithelium · local PDF available · normal oesophagus contains on average ~120 NOTCH1 mutations/cm²; more driver mutations in normal tissue than in oesophageal cancer itself. ↩

  6. cagan-2022-somatic-mutation-rates-mammals · doi:10.1038/s41586-022-04618-z · PMID 35418684 · n=208 crypts, 56 individuals, 16 species · cross-species comparative WGS · LME regression FVE=0.85, P=1×10⁻⁶ (lifespan correlation); slope=−0.86 (95% CI: −1.08, −0.65) · model: multi-mammalian intestinal crypts · local PDF available · 429 citations · FWCI 56.6 · 100th citation percentile · “Somatic mutation rates scale with lifespan across mammals” — foundational quantitative test of rate-lifespan prediction. ↩ ↩2

  7. albertson-2009-polymerase-proofreading-mice · doi:10.1073/pnas.0907147106 · PMID 19805137 · in-vivo mouse genetics · model: C57BL/6 Pol-ε/Pol-δ proofreading-deficient mice · mutator mice develop tissue-specific cancers (intestinal adenocarcinomas, thymic lymphomas) not accelerated systemic aging — the key complication for Prediction 2. ↩ ↩2

  8. doi:10.1038/s41586-025-09584-w · Lawson ARJ et al. · Nature 2025 · PMID 41062696 · n=1,042 oral epithelium + 371 blood samples · observational, population-scale, ultra-sensitive nanorate sequencing (targeted NanoSeq) · 14 citations · FWCI 89.0 · 100th citation percentile · local PDF available · verified against full PDF · “Somatic mutation and selection at population scale” — abstract states 46 genes under positive selection; body text (p. 3) reports 49 genes under positive selection by dNdScv (the difference reflects distinct sub-analyses); mutation rate in oral epithelium ~18 SNVs/diploid genome/year (slope 18.0, 95% CI 16.7–19.4); tobacco and alcohol exposures accelerate clonal dynamics; extends Martincorena programme to population scale across 1,042 individuals. ↩

  9. yang-2023-epigenetic-information-loss · doi:10.1016/j.cell.2022.12.027 · in-vivo (ICE mouse model + multi-tissue) · model: C57BL6/J mice · local PDF available · verified against full PDF · epigenetic aging rate “~50% faster in the ICE mice than Cre controls (p < 0.0001)” (Fig. 4K–M); OSK “reversed by up to 57%” DNAme age (Fig. 7G); no detectable difference in mutation frequency between Cre and ICE cells at I-PpoI recognition sites or genome-wide (WGS) — confirms aging driven by DSBs without permanent sequence change; primary evidence against somatic-mutation-theory causal primacy. ↩

  10. doi:10.1038/s41573-025-01212-6 · Bujarrabal-Dueso A, Garinis GA, Robbins PD, Vijg J, Schumacher B · Nat Rev Drug Discov 2025 · PMID 40506534 · review · 10 citations · FWCI 24.1 · 99th citation percentile · OA via PMC (PMC12825315) · PDF download failed (0 candidate URLs after filtering); abstract verified via PubMed efetch · “Targeting DNA damage in ageing: towards supercharging DNA repair” — states “DNA damage to the somatic genome is a major causal mechanism of ageing”; reviews pharmacological strategies to enhance repair capacity; consensus statement from leading DDT/somatic-mutation-theory proponents. no-fulltext-access (PMC URL identified but local download pipeline returned 0 candidate URLs — retry on next verification pass). ↩

Graph View

Backlinks