🧬 Aging Wiki

antagonistic-pleiotropy

Properties1
aliasesWilliams' theory, antagonistic pleiotropy hypothesis, AP hypothesis, Williams 1957

20 min read

Antagonistic Pleiotropy

The frame

This is a Mode B (conceptual-frame) page. Antagonistic pleiotropy organizes aging biology by explaining why natural selection permits aging to exist — it does not generate sharp, specific testable predictions at the molecular level. The frame should be read as an organizing scheme, not a mechanistic hypothesis.

The antagonistic pleiotropy (AP) hypothesis 1 proposes that natural selection cannot eliminate aging because some genes that confer fitness benefits early in life impose costs later in life, and early-life selection pressure is always stronger than late-life selection pressure. Once an organism has successfully reproduced, further survival contributes little to evolutionary fitness — Hamilton (1966) formalized this as the declining force of natural selection with age (the “selection shadow”) 2. Within this shadow, genes that were favored for early benefits persist even if their late-life costs are severe. Aging is therefore not an adaptation; it is the accumulated late-life cost of alleles optimized for early-life success.

The Williams 1957 paper introduced AP as a complement to Medawar’s (1952) mutation accumulation (MA) hypothesis. Whereas MA invokes the passive accumulation of late-onset deleterious mutations that selection never “sees” clearly enough to purge, AP posits an active mechanism: selection positively favors pleiotropic alleles that trade longevity for early reproductive advantage. The two hypotheses are complementary (both operate via the selection shadow) and are no longer considered mutually exclusive.

What it explains well

  • Why aging is essentially universal across sexually reproducing species. AP predicts aging wherever the force of natural selection declines with age — that is, in nearly all organisms with separate reproductive seasons or post-reproductive survival. The universality of aging is difficult to explain if aging were purely maladaptive drift; AP provides a positive-selection mechanism. See caenorhabditis-elegans, mus-musculus, drosophila-melanogaster for the empirical cross-species pattern (these pages may be stubs).

  • Why longevity-promoting interventions often impair early-life fitness in model organisms. Many genes whose loss-of-function extends lifespan — daf-2, age-1, chico, reduced mTOR — confer reduced growth, fecundity, or stress resistance in some early-life contexts. insulin-igf1 page documents the daf-2/Igf1r lifespan data; the fecundity trade-offs are noted there. This is the pattern AP predicts. unsourced — specific fecundity trade-off data per gene requires per-gene citation; link to individual protein/pathway pages.

  • The pleiotropic disease allele pattern in humans. Several human alleles that are maintained at high population frequency despite significant late-life disease costs can be interpreted through an AP lens — the assumption being they conferred early-life advantage that outweighed late costs under ancestral demography. APOE4 (elevated late-life Alzheimer’s and cardiovascular risk; possible early-life immune and reproductive benefits in pathogen-rich environments) and sickle-cell trait (heterozygote malaria resistance; homozygote sickle-cell disease) are canonical examples. Neither is unambiguously AP — the evidence base for early-life benefits of APOE4 is incomplete 3. needs-replication for APOE4 early-benefit claims.

  • The cancer-aging longevity tradeoff. Telomere biology offers a clean AP example: shorter telomeres suppress tumor formation (beneficial early, especially in species with large bodies and many cell divisions) but accelerate cellular senescence and stem-cell exhaustion with age. Longer telomeres slow aging but increase cancer risk. telomere-attrition (hallmark page) documents the telomere-aging axis; the cancer-aging trade-off interpretation is consistent with AP. See also cellular-senescence for the p53-driven tumor-suppression / senescence-burden duality.

  • p53 as a molecular AP example. The tumor-suppressor protein p53 is the canonical molecular illustration of AP 4: elevated p53 activity suppresses cancer (powerful early-life benefit for multicellular organisms) but promotes cellular senescence, tissue atrophy, and accelerated aging phenotypes when constitutively hyperactive. Tyner et al. 2002 demonstrated this trade-off using p53+/m mice (carrying an m-allele that deletes exons 1–6 and expresses a C-terminal p53 fragment conferring constitutive activation): p53+/m mice had <6% tumour incidence vs >45% in wild-type littermates but a median lifespan of 96 weeks vs 118 weeks for p53+/+ — plus pronounced aging phenotypes (lordokyphosis, osteoporosis, muscle atrophy, reduced organ mass). A second pL53 temperature-sensitive allele line also exhibited early aging phenotypes, confirming the pattern across two independent p53 perturbations. See p53-pathway (verified) for the molecular detail.

  • mTOR and IGF-1 as systemic AP examples. mTOR and insulin/IGF-1 signaling drive growth, anabolism, and reproductive function early in life — functions under strong positive selection. The same pathways promote cancer, cellular hypertrophy, and loss of proteostasis in aged tissues. See mtor (verified) and insulin-igf1 (verified) for molecular and lifespan data.

  • The dog breed-size lifespan paradox — within-species artificial-selection demonstration of the somatotropic-axis trade-off. Across mammalian species, larger body size correlates with longer lifespan; within the domestic dog, the relationship is inverted — giant breeds (Great Danes, Mastiffs) median ~6–8 yr, small breeds (Chihuahuas, Yorkies) 14–18 yr. Sutter et al. 2007 identified an IGF1 haplotype as a major determinant of breed body size 5, and Kraus 2013 showed that the dominant explanation for the lifespan difference is accelerated aging in large breeds rather than higher extrinsic mortality 6. This is a within-species replication of the same somatotropic-axis-vs-aging trade-off documented at the mTOR/IGF-1 systemic level — at single-haplotype molecular resolution. Important caveat: dog breeds are products of artificial selection, not natural selection — humans selected for body size and the longevity cost was a byproduct, not an evolutionarily fixed AP allele. This makes the dog example a useful mechanistic demonstration of the somatotropic-axis trade-off but not a textbook natural-selection AP case. The Loyal Inc. LOY-001/003 program (FDA RXE 2023 for LOY-001) attempts to reverse this within-species AP by reducing circulating IGF-1 in large breeds to small-breed levels. See canis-lupus-familiaris for the full breed-size-lifespan biology and active research-program context.

  • Acute inflammation vs. chronic inflammaging. Innate immune activation (NF-ÎşB axis, IL-6, TNF-α) provides critical early defense against pathogens — high selection value. The same inflammatory circuitry, chronically activated in aged tissue, drives inflammaging and multi-system pathology. This is not a single-gene AP effect but a systems-level trade-off consistent with the frame. unsourced — link to chronic-inflammation when page exists.

  • Sanchez 2026 — disease-tolerance genes show explicit antagonistic pleiotropy. A 2026 Nature paper (Ayres lab, Salk) directly demonstrated AP at single-gene resolution in a polymicrobial sepsis mouse model: in young mice, cardiac Foxo1 and its downstream effector Trim63 (MuRF1) protected against sepsis-induced cardiac remodeling, multi-organ injury, and death; in aged mice, the same genes acted as drivers of sepsis pathogenesis and mortality 7. Same gene, opposite effect on survival across the lifespan — this is the experimentally cleanest demonstration of AP at the molecular level published since Tyner 2002 p53. The authors explicitly frame this as evidence that disease-tolerance genes show antagonistic pleiotropy. The finding has direct therapeutic implications: stratifying immunomodulatory therapy by host age, since the same intervention may be protective or harmful depending on whether Foxo1/Trim63 are in their early-life or late-life regime.

  • Kobayashi 2026 — reproductive endometrial AP via mTORC1. A 2026 Reproduction review integrates reproductive-aging biology with the AP frame, arguing that physiological mTORC1 activation supports stromal cell proliferation and endometrial decidualization in young women but, when sustained in aging or metabolic dysfunction, drives endometrial cellular senescence and SASP-mediated implantation failure 8. This is a tissue-specific AP example mechanistically linking mtor-axis hyperfunction to age-specific reproductive outcomes.

  • cGAS as a molecular AP exemplar — antiviral defense vs. HR suppression and inflammaging (Chen 2025). cgas evolved primarily as a cytosolic DNA sensor: free cytoplasmic DNA (from viral infection, mitochondrial leakage, or micronuclei) binds cGAS → cGAMP synthesis → STING activation → type I IFN response. This antiviral function confers a clear early-life survival benefit. However, cGAS also localizes to the nucleus, where — in humans and mice — it suppresses homologous-recombination (HR) repair 9. The mechanistic handle is the chromatin-eviction cycle: trim41 (E3 ubiquitin ligase) ubiquitinates nuclear cGAS at damage sites, marking it for extraction by the VCP segregase, which physically strips cGAS from chromatin and relieves the HR obstruction. This recovery mechanism means the HR-suppressive cost is transient per damage event, but over a lifetime of DNA damage events the cumulative effect is meaningful: chronically dampened HR fidelity contributes to genomic-instability. A second late-life cost compounds the first: in aged tissues, cytosolic cGAS is persistently over-activated by chromatin fragments from dying or senescent cells (CCFs) and leaking mitochondrial DNA, driving inflammaging via the STING–NF-ÎşB axis.

    This is a strong AP exemplar because the early-life benefit (antiviral IFN defense) and the two late-life costs (HR suppression + inflammaging) are mechanistically unified through the same protein’s dual-compartment activity: the chromatin-residency/eviction cycle links the nuclear HR-suppressive arm to the cytosolic sensing arm, and chronic activation of the sensing arm in aged tissue drives systemic inflammation. The molecular handle is specific and potentially tractable.

    Comparative biology natural experiment (Chen 2025). The naked mole-rat (heterocephalus-glaber), the longest-lived rodent (~30 yr vs ~3 yr for mice), has independently tuned this trade-off. NMR cGAS carries a four amino acid divergence that weakens TRIM41-mediated ubiquitination and P97/VCP segregase interaction — making chromatin eviction less efficient. Consequently, NMR cGAS undergoes prolonged chromatin retention after DNA damage. Rather than being HR-suppressive, the retained NMR cGAS acts as a scaffold that enhances FANCI–RAD50 interaction at damage sites, potentiating MRN complex assembly and HR repair. The same four residues are sufficient to mediate cGAS’s function in antagonizing cellular and tissue aging, and their introduction is associated with lifespan extension in experimental systems 10. The NMR thus represents a natural experiment in which evolution shifted the TRIM41–P97–cGAS axis toward the late-life genomic-stability benefit at the expense of more rapid chromatin eviction.

    Open question — does the NMR pay an antiviral cost? Prolonged nuclear sequestration of cGAS could in principle reduce the cytosolic cGAS pool available for rapid IFN response to viral challenge. Whether NMR has compensatory antiviral mechanisms or an altered antiviral phenotype is not addressed in the Chen 2025 abstract. This is the critical open-question for the AP framing: if NMR cGAS pays no antiviral cost, the trade-off has been escaped rather than merely shifted. If it does pay a cost, this is a textbook AP case — reduced late-life genomic instability at the price of reduced early-life antiviral capacity. no-fulltext-access — Chen 2025 full text needed; NMR antiviral phenotype not characterized in abstract.

    See chen-2025-nmr-cgas-hr-repair (abstract-only, not_oa) and liu-2018-nuclear-cgas-hr-suppression for the prior-art nuclear-cGAS suppression result.

2026 theoretical refinements and tensions

  • Bega & Hadany 2026 — accelerating-damage extension to Hamilton. A 2026 Ecol Evol paper expands Hamilton’s classic selection-force-with-age model to allow gene effects on mortality to accelerate with age throughout the lifespan, rather than being confined to a single age-of-onset 11. Their model shows that under biological constraints on reducing internal damage, a positive feedback loop can develop where senescence-slowing genes increase selection for further senescence retardation — providing a theoretical mechanism for negligible senescence within the AP/MA framework. The paper is framed as a friendly extension of AP, not a refutation; it preserves the selection-shadow logic but allows richer dynamics. This is the most substantive theoretical refinement of AP since Williams 1957.

  • Ringel 2025 — optimization theories vs weakening-selection theories. A 2025 Biol Rev (Cambridge) review (Ringel, Life Biosciences) classifies aging theories into three categories — mechanistic, weakening-force-of-selection (which AP and MA both belong to), and optimization theories (which posit aging is positively selected under resource constraints) — and argues that only optimization theories fit the full empirical dataset (CR effects, long-lived organisms, mortality plateaus, eusocial queen longevity, lifespan malleability) 12. The author predicts that lifespan-extending mutations should generally reduce fitness under natural conditions and notes that available evidence supports this prediction. This is a strong industry-aligned (Life Biosciences) push to reframe AP and disposable-soma as special cases of a broader optimization-theory category. Status implication: if Ringel’s framework gains traction, AP may be reclassified as a subset of an optimization-theory family rather than an independent active-frame. The exact taxonomic relationship remains contested.

  • Gems 2025 — multifactorial two-stage AP. A 2025 Aging (Albany NY) synthesis (Gems lab, UCL) proposes that aging proceeds in two stages: first, accumulation of latent injury foci (infection, mechanical injury, somatic + inherited mutation); second, late-life “germination” of these foci via wild-type gene action including antagonistic pleiotropy 13. The framework treats AP as one of several “etiologies largely confined to aging” rather than the singular driver — consistent with the polygenic-AP-landscape modern interpretation but elevating other causes (infection latency, somatic mutation) to comparable causal weight. This is a softening of single-mechanism AP toward a multifactorial pluralism.

What it doesn’t easily address

  • Specific molecular predictions. AP does not tell you which genes are pleiotropic, at what allele frequencies, or what the magnitude of the early-benefit / late-cost trade-off must be to fix in a population. The frame predicts the existence of such alleles; it does not derive their identity. Austad & Hoffman (2018) note that while AP effects have been found in nearly every case where they were seriously investigated in the laboratory, the actual AP alleles remain largely undiscovered in natural populations, and studies reporting lifespan extension often fail to report early fitness costs at all 3. needs-replication

  • The just-so risk. AP is readily invoked post hoc to explain almost any harmful late-life gene effect by hypothesizing an early-life benefit that is difficult to measure directly (especially under ancestral demographic conditions). Austad & Hoffman (2018) flag this concern: many studies of lifespan-extending mutations do not report reproductive or other early fitness components, making it difficult to establish that the allele was truly positively selected rather than merely tolerated 3. The frame is not falsifiable at the level of individual alleles unless both early benefits AND late costs are independently measured. no-mechanism for many specific alleged AP genes.

  • Post-reproductive life. AP predicts negligible selection on post-reproductive phenotypes. But human post-reproductive lifespan is substantial (decades), and the “grandmother hypothesis” posits that post-reproductive females contribute to inclusive fitness via kin assistance. The frame provides no clear account of what drives aging within the post-reproductive period, once alleles with early benefits have already exerted their effects.

  • Species-specific longevity variation. While AP explains universal aging, it does not easily explain why some species live 2 years and others 500. For within-clade longevity variation, the disposable-soma frame (extrinsic hazard driving optimal somatic maintenance budget) does more explanatory work. See disposable-soma-theory (verified).

  • Distinguishing AP from mutation accumulation empirically. Both mechanisms predict accumulating late-life pathology in the selection shadow. Separating the two requires identifying whether late-acting alleles were positively selected (AP) or merely tolerated due to weak purifying selection (MA) — a distinction requiring evolutionary genetic data that is rarely available for aging-relevant human loci.

Where specific predictions get tested

AP-derived predictions are tested at the level of specific genes and pathways, not at the level of the frame itself. The relevant atomic pages where AP predictions are implicitly tested include:

  • p53-pathway (verified) — p53 hyperactivation trade-off: tumor suppression vs accelerated aging
  • mtor (verified) — mTOR inhibition extends lifespan but impairs anabolic programs
  • insulin-igf1 (verified) — Igf1r/daf-2 loss-of-function lifespan vs. early fitness costs
  • telomere-attrition — telomere length / cancer trade-off (page status: check)
  • cellular-senescence — senescence as tumor suppressor vs. SASP-driver trade-off
  • disposable-soma-theory (verified) — the companion evolutionary frame that makes the resource-allocation version of these predictions quantitative

  • disposable-soma-theory (verified) — Kirkwood 1977 14; complementary frame; where AP operates at the gene-selection level, disposable soma operates at the metabolic resource-allocation level. Kirkwood himself acknowledged Williams’ AP hypothesis in the disposable-soma paper and treated them as complementary. The two frames together answer: why does natural selection permit aging (AP) and what determines the rate of aging across species (DS).

  • free-radical-theory-of-aging (verified, superseded) — Harman 1956; a mechanistic hypothesis rather than an evolutionary frame; AP provides an evolutionary reason why antioxidant investment might be sub-optimal (DS/AP frame predicts it would be), but AP does not make specific ROS claims.

  • information-theory-of-aging (verified) — Sinclair 2013+; the epigenetic information loss hypothesis is mechanistic; AP provides a reason why epigenome maintenance machinery might be evolutionarily sub-optimal (selection pressure declines post-reproduction), but the two operate at different levels of description.

  • Mutation accumulation (Medawar 1952) — book chapter; no DOI. Complementary evolutionary mechanism; AP and MA both rely on the declining force of natural selection with age; differ in whether late-acting alleles were positively selected (AP) or merely not purged (MA). Often treated as jointly operating. unsourced — no dedicated wiki page yet for mutation accumulation.

Notes / why this frame matters

  1. AP as an intervention-design guide. If aging-promoting alleles were fixed because they conferred early-life benefit, late-onset interventions (after reproduction is complete, or targeting only the late-life arm of the trade-off) can in principle be deployed without disrupting the early-life functions. This is the evolutionary argument for why rapamycin started in late-life mice avoids the growth-suppression trade-off, and why senolytic targeting of post-mitotic senescent cells may not impair development. The frame is useful for deciding when to intervene.

  2. The polygenic nature of AP in practice. The original Williams 1957 framing imagined relatively discrete pleiotropic alleles with identifiable early benefits and late costs. The modern interpretation — informed by GWAS and polygenic score analyses — is that age-related disease architectures are highly polygenic and the early/late fitness effects of individual SNPs are too small to measure directly. The “AP landscape” is thus more a diffuse background selection pressure than a set of identifiable AP genes 3.

  3. AP is not the same as trade-off. Not every aging/reproduction trade-off is AP. Disposable soma is a resource-allocation trade-off, not necessarily AP. AP specifically requires: (a) a single gene or allele, (b) positive fitness effect early in life, (c) negative fitness effect late in life, (d) net positive selection because early fitness outweighs late cost. Many observed aging-reproduction trade-offs are physiological or resource-based rather than gene-level pleiotropic. Conflating the two is common in popular accounts.

  4. The “active-frame” status. AP is not falsifiable in the strict Popperian sense — it is a framework for interpreting findings, not a mechanistic claim with falsifiable predictions. Its value lies in organizing which gene-level results count as AP examples, guiding intervention timing, and motivating the search for early-life benefits of disease alleles. It is not expected to be “confirmed” or “falsified” as a whole; individual AP examples can be challenged or corroborated, and that process continues. Austad & Hoffman 2018 3 conclude that AP is “somewhere between very common or ubiquitous throughout the animal world” and that late-life medical interventions do not have to be deployed early (after reproduction the early-life fitness trade-offs are moot), which is an optimistic implication for aging medicine.

Limitations and gaps

no-fulltext-access — Williams 1957 (doi:10.1111/j.1558-5646.1957.tb02911.x), Hamilton 1966 (doi:10.1016/0022-5193(66)90184-6), and Kirkwood 1977 (doi:10.1038/270301a0) are closed-access; no local PDF. Claims attributed to these sources are based on widely-cited secondary sources and the DOI lookup titles. Verify before relying on exact phrasings. Note: Austad & Hoffman 2018 (doi:10.1093/emph/eoy033) PDF downloaded and verified 2026-05-06. needs-replication — APOE4 early-life benefit claims (specific mechanism not established). unsourced — mutation accumulation (Medawar 1952) frame has no dedicated wiki page yet; cross-reference is to a book with no DOI. unsourced — inflammaging as systemic AP example needs dedicated chronic-inflammation page.

Footnotes

  1. doi:10.1111/j.1558-5646.1957.tb02911.x · Williams GC · Evolution 1957 · “PLEIOTROPY, NATURAL SELECTION, AND THE EVOLUTION OF SENESCENCE” · n/a (theoretical) · foundational AP paper; 4,226 citations per OpenAlex; closed-access no-fulltext-access ↩

  2. doi:10.1016/0022-5193(66)90184-6 · Hamilton WD · Journal of Theoretical Biology 1966 · “The moulding of senescence by natural selection” · n/a (theoretical / mathematical) · formalized the declining force of selection with age (“selection shadow”); 2,279 citations per OpenAlex; closed-access no-fulltext-access ↩

  3. doi:10.1093/emph/eoy033 · Austad SN, Hoffman JM · Evolution Medicine and Public Health 2018 · “Is antagonistic pleiotropy ubiquitous in aging biology?” · review · 183 citations; gold OA; local PDF available at DOI lookup · conclusion: AP is “somewhere between very common or ubiquitous throughout the animal world”; notes that many lab lifespan-extension studies do not report early fitness costs; trade-offs are consistent with but rarely proven to be AP (vs. disposable soma) in natural populations ↩ ↩2 ↩3 ↩4 ↩5

  4. doi:10.1038/415045a · Tyner SD et al. · Nature 2002 · “p53 mutant mice that display early ageing-associated phenotypes” · model: Mus musculus (p53+/m allele: exons 1–6 deletion expressing C-terminal fragment; also pL53 temperature-sensitive allele) · canonical molecular AP illustration; p53+/m median lifespan 96 wk vs 118 wk p53+/+; <6% tumour vs >45% tumour; 1,434 citations ↩

  5. doi:10.1126/science.1137045 · Sutter NB, Bustamante CD, Chase K et al. · Science 2007 · “A single IGF1 allele is a major determinant of small size in dogs” · observational · model: Canis lupus familiaris across multiple breeds · identified IGF1 haplotype as major determinant of breed body size · needs-citation-verification — DOI not yet verified against PDF in this propagation pass; verifier on [[model-organisms/canis-lupus-familiaris]] will confirm ↩

  6. doi:10.1086/669665 · Kraus C, Pavard S, Promislow DEL · American Naturalist 2013 · “The size–life span trade-off decomposed: why large dogs die young” · observational · model: dog breed lifespan dataset (~74 breeds) · demonstrated that within-dog body-size lifespan inversion is primarily driven by accelerated aging rate, not by elevated extrinsic mortality · needs-citation-verification — DOI not yet verified against PDF in this propagation pass; verifier on [[model-organisms/canis-lupus-familiaris]] will confirm ↩

  7. doi:10.1038/s41586-025-09923-x · Sanchez KK, McCarville JL, Stengel SJ, Snyder JM, Williams AE, Ayres JS · Nature 2026;650(8102):727-735 · in-vivo mouse · polymicrobial sepsis (LD50) in young vs aged C57BL/6 · cardiac Foxo1 + Trim63 (MuRF1) protective in young (anti-cardiac-remodeling) but pathology-driving in aged — single-gene-resolution antagonistic-pleiotropy demonstration · OA (Salk/HHMI); PMC12916300; PMID 41535469 · verified-scope: PubMed efetch abstract only · cleanest molecular AP demonstration since Tyner 2002 p53 ↩

  8. doi:10.1093/reprod/xaag044 · Kobayashi H, Nishio M, Umetani M, Shigetomi H, Imanaka S, Hashimoto H · Reproduction 2026;xaag044 · review · mTORC1 hyperactivation as endometrial-receptivity AP example; physiological in young, drives senescence/SASP/implantation failure in aging · PMID 41965078 · verified-scope: PubMed efetch abstract only ↩

  9. liu-2018-nuclear-cgas-hr-suppression · n=NR · in-vitro+in-vivo · doi:10.1038/s41586-018-0629-6 · PMID:30356214 · Liu H et al. · Nature · 2018 · model: human cell lines + mouse xenograft · “Nuclear cGAS suppresses DNA repair and promotes tumorigenesis” · established that nuclear cGAS inhibits HR repair in human and mouse cells via chromatin retention; key prior-art establishing the cGAS dual-compartment biology that Chen 2025 extends to the NMR comparative system · closed-access; not in a local paper archive · no-fulltext-access ↩

  10. chen-2025-nmr-cgas-hr-repair · Science 2025 · doi:10.1126/science.adp5056 · PMID 41066557 · in-vitro+in-vivo · multiple organisms (NMR, mouse, cell lines) · NMR cGAS 4-AA divergence weakens TRIM41 ubiquitination + P97 segregase eviction → prolonged chromatin retention → enhanced FANCI–RAD50 interaction → HR repair potentiation → delayed aging and extended lifespan in experimental systems · closed-access (not_oa); abstract-only · #gap/no-fulltext-access ↩

  11. doi:10.1002/ece3.72988 · Bega D, Hadany L · Ecol Evol 2026;16(3):e72988 · theoretical · extends Hamilton’s classic selection-force model with age-accelerating damage; shows positive feedback loop in senescence retardation; sheds light on Peto’s paradox, Strehler-Mildvan, and negligible senescence · OA gold; PMC12936436; PMID 41766736 · verified-scope: PubMed efetch abstract only ↩

  12. doi:10.1002/brv.70109 · Ringel MS · Biol Rev Camb Philos Soc 2026;101(2):911-925 · review · “Why we age” — categorizes aging theories into mechanistic / weakening-selection / optimization; argues optimization theories (which subsume AP/MA/disposable-soma) fit the broadest empirical dataset; predicts lifespan-extending mutations generally reduce natural-condition fitness · OA gold; PMC12965854; PMID 41366834 · verified-scope: PubMed efetch abstract only · author affiliated with Life Biosciences ↩

  13. doi:10.18632/aging.206339 · Gems D, Carver A, Zhao Y · Aging (Albany NY) 2025;17(12):2989-3002 · review/theory · proposes two-stage aging: latent injury accumulation + late-life “germination” via wild-type gene action including AP; multifactorial pluralism softening single-mechanism AP · OA; PMC13147727; PMID 41474639 · verified-scope: PubMed efetch abstract only · UCL Institute of Healthy Ageing ↩

  14. doi:10.1038/270301a0 · Kirkwood TBL · Nature 1977 · “Evolution of ageing” · n/a (theoretical) · disposable soma theory; companion evolutionary frame to AP; 1,867 citations per OpenAlex; closed-access no-fulltext-access ↩

Graph View

Backlinks