A brief history of the hyperfunction theory of aging and future directions (paper July 24 2026)

John_Hemming · July 25, 2026, 8:22am

https://www.aging-us.com/article/206403/text

chatGPT(5.5paid):

As people know I think the biochemistry of ageing and development at the core relate to the same pathway which changes gene expression through a mixture of reduced mRNA synthesis, changes in splicing (including NMD splices) and reductions in expressing some genes. That is close to hyperfunction as defined by Blagoskonny, but not the same.

Paper overview

Title: A brief history of the hyperfunction theory of aging and future directions
Author: João Pedro de Magalhães
Article type: Research Perspective, published in Aging in 2026.

This is primarily a historical and conceptual perspective, not an experimental study. It traces the development of programmatic theories of ageing, highlights Mikhail Blagosklonny’s hyperfunction theory, and argues that ageing may arise substantially from developmental and growth programmes that continue beyond their useful period.

Summary

Central argument

The paper contrasts two broad views of ageing:

Damage-based theories, in which ageing results mainly from the gradual accumulation of molecular errors and damage.
Programmatic theories, in which normal developmental, growth and regulatory programmes persist or become dysregulated in later life, producing pathology.

The author carefully distinguishes programmatic ageing from programmed ageing. Programmed ageing implies that ageing itself evolved as an adaptive process. Programmatic theories instead propose that ageing is a non-adaptive consequence of biological programmes selected for their benefits earlier in life.

The hyperfunction theory is presented as a mechanistic form of antagonistic pleiotropy: pathways that promote growth, development and reproduction early in life remain excessively active later, producing cellular and tissue hyperfunction, disease and eventual functional decline.

Historical development

The paper traces relevant ideas from:

Weismann’s early proposals concerning programmed ageing.
McCay’s caloric-restriction experiments, originally related to the relationship between growth and lifespan.
Williams’s antagonistic pleiotropy theory.
Kirkwood’s disposable soma theory.
Later developmental and quasi-programme theories.

The author describes how damage-centred explanations became dominant during the growth of molecular biology, because age-associated defects could be identified in almost every cellular process.

He then recounts evidence that led him away from simple damage-accumulation models, including:

Increased oxidative damage in mice with reduced mitochondrial superoxide dismutase without shortened lifespan.
Weak relationships between metabolic rate and mammalian longevity.
The indefinite proliferation of telomerase-expressing human cells.
Strong relationships among species-specific development rates, growth rates and lifespan.
The ability of single-gene manipulations, particularly in the GH–IGF-1 pathway, to substantially alter lifespan.

Blagosklonny and hyperfunction

The paper credits Blagosklonny with crystallising the idea of ageing as a quasi-programme: developmental programmes continue running after development is complete, becoming excessive and harmful.

In this framework:

mTOR promotes growth and development early in life.
Continued mTOR activity later contributes to hypertrophy, cellular hyperfunction, senescence and age-related disease.
Rapamycin slows these processes by inhibiting mTOR.

The author regards rapamycin’s reproducible extension of lifespan in mice as important evidence in favour of the theory, although he argues that mTOR alone is unlikely to explain major species differences in growth, development and ageing.

Author’s extension of the theory

De Magalhães relates hyperfunction to his own “software design flaw” hypothesis. This treats the genome and epigenome as an informational architecture optimised for development and reproduction, but not for indefinite maintenance.

He proposes that:

Development progressively reduces cellular plasticity and regenerative potential.
These changes may initially reduce cancer risk.
Continued suppression of proliferation and plasticity later impairs tissue repair.
Ageing may therefore reflect persistent regulatory states established during development.

The paper presents cancer and ageing as partly opposing outcomes: mechanisms that constrain proliferation may protect against cancer early but contribute to degeneration later.

Future research directions

The author argues for greater attention to the entire life course rather than studying only adult and old organisms. Suggested research priorities include:

Mapping developmental gene-regulatory trajectories into old age.
Testing whether repair and maintenance systems are deliberately downregulated during development.
Investigating pathways such as Wnt, Hedgehog and Hippo.
Identifying tissue-specific rejuvenation factors.
Distinguishing regulatory resetting from conventional molecular-damage repair.
Testing partial reprogramming and related interventions in vivo.

The paper ends with a hybrid position: ageing probably includes both programmatic and damage-related processes, but the author suspects that programmatic mechanisms account for much of non-cancer degenerative ageing.

Novelty

The novelty is mainly conceptual synthesis and historical framing, rather than new empirical evidence.

1. Integration of several related theories

The paper brings together:

Hyperfunction theory.
Antagonistic pleiotropy.
Developmental run-on.
The software design flaw hypothesis.
Cancer–ageing trade-offs.
Epigenetic reprogramming.

This creates a relatively unified account in which ageing arises from regulatory programmes that were selected for development, reproduction and early-life cancer suppression, but become maladaptive later.

That integration is useful because these ideas are often discussed separately.

2. First-person historical account

The author provides a personal account of the parallel development of his ideas and Blagosklonny’s. This includes their scientific correspondence, areas of agreement and their differing emphases:

Blagosklonny: mTOR, pathology and clinical implications.
De Magalhães: comparative biology, genomics, epigenetics and species differences.

This material adds value as an intellectual history, particularly in documenting how programmatic ageing theories developed during the early 2000s.

3. Explicit falsifiability criteria

A particularly useful section proposes observations that would substantially weaken the hyperfunction theory, including:

A two- to threefold lifespan extension through reduction of molecular errors alone.
Broad multi-organ ageing retardation through damage reduction without altering developmental signalling.
Rejuvenation that can be fully explained by molecular repair rather than regulatory resetting.

Many conceptual ageing papers do not clearly state what evidence would falsify their preferred theory, so this is a strength.

4. Developmental framing of rejuvenation

The paper argues that partial reprogramming should be interpreted not merely as damage repair, but as evidence that age-related states may be encoded in reversible gene-regulatory programmes.

It further proposes that tissue-specific factors capable of restoring a youthful rather than embryonic state may be more useful than broad Yamanaka-factor reprogramming.

5. Cancer suppression as a quasi-programme

The suggestion that declining proliferation, plasticity and regenerative capacity may begin as adaptive cancer-protective mechanisms and later contribute to ageing is an interesting extension of antagonistic pleiotropy.

However, this is largely an elaboration of the author’s recent prior work rather than an entirely new hypothesis introduced in this paper.

Critique

Strengths

Clear conceptual distinction

The distinction between programmed and programmatic ageing is important and well explained. It prevents a common misunderstanding that developmental theories necessarily imply that ageing evolved for the benefit of the organism or species.

Strong historical narrative

The paper gives a readable account of how ageing theory moved from developmental ideas to damage-centred theories and then partly back toward gene-regulatory explanations.

It also appropriately situates hyperfunction within evolutionary theory rather than presenting it as an alternative to evolution.

Recognition of a hybrid model

Although the paper strongly favours programmatic mechanisms, it does not completely deny damage. The conclusion that ageing is likely a hybrid phenotype is more defensible than an absolute claim that damage is irrelevant.

Translational relevance

The theory generates practical predictions:

Interventions targeting upstream regulatory programmes should have broader effects than interventions repairing individual forms of damage.
Lifelong developmental trajectories may reveal intervention targets.
Rejuvenation may require regulatory resetting rather than removal of accumulated lesions.

These are useful hypotheses for experimental design.

Main weaknesses

1. Evidence for genetic control does not establish a programme

One of the paper’s recurring arguments is that because single genes, GH–IGF-1 signalling, caloric restriction and rapamycin can alter lifespan, ageing must be substantially programmatic.

This inference is not conclusive.

A genetically regulated process is not necessarily a developmental quasi-programme. Genes can influence:

Damage production.
DNA repair.
Proteostasis.
Antioxidant capacity.
Immune function.
Metabolic stress.
Resistance to environmental injury.

For example, reducing mTOR could extend lifespan because it lowers protein synthesis, increases autophagy and reduces metabolic damage. Those observations are compatible with hyperfunction theory, but they are also compatible with modified damage-maintenance models.

The paper therefore sometimes treats evidence for modifiability as evidence for a specific causal architecture.

2. Rapamycin is not uniquely diagnostic of hyperfunction

Rapamycin is presented as major empirical support for the theory. It is indeed consistent with it, but rapamycin has many effects:

Reduced translation.
Increased autophagy.
Altered mitochondrial metabolism.
Reduced inflammation.
Changes in stem-cell function.
Improved proteostasis.
Reduced cancer incidence.

Because these effects span both regulatory and damage-control mechanisms, rapamycin cannot by itself distinguish hyperfunction from competing models.

A stronger test would require showing that lifespan extension depends specifically on correcting inappropriate developmental signalling rather than on improving damage clearance or stress resistance.

3. The critique of damage theories is somewhat selective

The paper emphasises findings that challenge simple free-radical or rate-of-living theories. Those criticisms are valid, but modern damage theories are broader and more sophisticated than the original free-radical theory.

Failure of oxidative damage alone to determine lifespan does not disprove important roles for:

Somatic mutation.
Mitochondrial DNA alterations.
Protein aggregation.
Extracellular-matrix crosslinking.
Clonal haematopoiesis.
Epigenetic instability.
Irreversible cell loss.

The paper sometimes moves too quickly from “oxidative damage is not the sole driver” to “damage is probably secondary.”

4. Cellular immortality is a weak analogy for organismal ageing

The observation that telomerase-expressing human cells can proliferate for extended periods in culture is used to question unavoidable damage accumulation.

However, immortalised cell populations do not demonstrate indefinite preservation of a complex organism. Cultures may continually select fitter clones, dilute damaged components through division and operate without the architecture and functional demands of organs.

Moreover, indefinite proliferation is not equivalent to indefinite maintenance of differentiated function.

The argument is suggestive, but it is not strong evidence against organism-level damage accumulation.

5. “Hyperfunction” can become too elastic

The theory risks explaining both excessive activity and loss of function by placing loss of function downstream of earlier hyperfunction.

For example:

Excessive growth signalling is hyperfunction.
Senescence is caused by hyperfunction.
Tissue degeneration is caused by senescent-cell hyperfunction.
Reduced regeneration is caused by developmental programmes continuing or by programmes being downregulated.

This flexibility makes the theory difficult to falsify unless specific pathways, time courses and tissue-level predictions are defined in advance.

The paper acknowledges that the theory remains abstract, but it does not fully solve this problem.

6. Limited disease-level validation

The author notes that presbyopia, thymic involution and perhaps osteoarthritis may fit a programmatic model. However, these examples remain relatively few.

For major age-associated conditions—including neurodegeneration, sarcopenia, chronic kidney disease, fibrosis and vascular calcification—the paper does not provide detailed causal chains showing:

The developmental programme involved.
Its normal early-life function.
Its continued or mistimed activity.
The molecular transition to pathology.
Evidence that selectively switching it off prevents disease.

Without such examples, the theory remains more a broad interpretative framework than a mature mechanistic model.

7. Species differences are raised but not mechanistically resolved

The author correctly argues that mTOR alone is unlikely to explain why mice and humans have radically different developmental and ageing rates.

However, the paper does not offer a sufficiently detailed alternative. References to the genome, epigenome and developmental timing are plausible but broad.

A stronger model would specify measurable parameters such as:

Rates of developmental transcriptional change.
Chromatin-state stability.
Tissue-specific enhancer turnover.
Stem-cell programme shutdown.
Growth-pathway pulse duration.
Scaling of repair-system expression across species.

8. Partial reprogramming is not yet decisive evidence

Partial reprogramming is presented as potentially strong support for programmatic ageing. Yet improvement after reprogramming could reflect multiple effects:

Restoration of gene-regulatory patterns.
Enhanced DNA repair.
Improved mitochondrial quality control.
Removal or dilution of damaged proteins.
Selection against damaged cells.
Altered inflammatory signalling.

To establish the programmatic model, studies would need to show that rejuvenation occurs even when accumulated lesions remain largely unchanged, or that resetting a defined regulatory state is sufficient to restore function.

9. The article has an advocacy tone

The paper is partly a tribute to Blagosklonny and explicitly states that he may be proven increasingly correct. That is appropriate for a perspective or commemorative issue, but it creates a risk of confirmation bias.

Competing theories receive less sympathetic treatment and there is limited discussion of evidence that is difficult for hyperfunction theory to explain.

The declared commercial interests in rejuvenation gene therapy are relevant, particularly because the paper presents partial reprogramming as an important future direction. This does not invalidate the argument, but it should encourage careful separation of evidence from speculation.

Overall assessment

This is a valuable, readable and intellectually stimulating perspective. Its strongest contribution is not proving the hyperfunction theory, but articulating a research programme in which ageing is studied as a continuation or misregulation of development rather than simply as accumulated wear.

The paper succeeds in showing that:

Ageing cannot be adequately explained by simplistic oxidative-damage models.
Growth and nutrient-sensing pathways have major effects on lifespan.
Developmental timing and lifespan are closely connected.
Regulatory states associated with ageing may be at least partly reversible.

It does not, however, establish that hyperfunction is the dominant cause of ageing. Much of the cited evidence is equally compatible with hybrid models in which regulatory programmes influence the production, repair, tolerance and clearance of molecular damage.

The most defensible interpretation is therefore:

Ageing is likely generated by interactions between persistent or mistimed gene-regulatory programmes and accumulating molecular and structural damage. Hyperfunction theory provides a useful account of the upstream regulation, but it has not yet demonstrated that downstream damage is mostly incidental rather than causally important.

The paper’s greatest future value will depend on converting the general concept of developmental “run-on” into precise, tissue-specific, experimentally falsifiable pathways.