Battle of Ideas

Aging is a disease we should treat

A steelman map of the public fight over aging disease — arguments for and against, with sources and assumptions.

AI-generated · paired steelman agents · independently red-teamed · Pass-1 source spot-checks only · framing-fidelity not independently verified · single model family

Whether aging itself is a medical indication to treat (geroscience, longevity drugs), not whether old people deserve care. ICD coding, 'natural', and who gets the extra years are in the fight. Not a twin of the live free-will page.

AGAINST 5

no further strong arguments at this depth

FOR 5

no further strong arguments at this depth

People also ask

Questions people actually type. The two columns above are the cases — open a card for sources, assumptions, and counters.

Is aging a disease?

Should we treat aging as a disease?

Longevity medicine

Can aging be treated?

Why aging is not a disease

Treat aging as disease

Ordering within each column: strongest first — validation tier, then source quality, then representativeness.

AGAINST · Aging is a disease we should treat
Empirical — moderateP1

We already treat the diseases of old age — that is geriatrics, not a new kingdom

The extra years rich countries already bought were blood pressure pills, statins, stents, screening, and ART — disease programmes. Compression of morbidity, when it happens, happens through those. Geroscience says one hammer on the hallmarks beats many hammers on the wreckage. AGAINST's reply: show the human hammer. Rapamycin in heterogeneous mice (Harrison, Nature, 2009) is a real animal result. Metformin's observational aging story is entangled with diabetes. TAME is uncompleted. Sinclair's public stack and de Grey's repair schematic are the marketing department the scope told this page not to take as gospel. Until a trial shows that treating 'aging' does something disease-by-disease care does not, the indication is a renaming. Renaming has a cost: research money, prescription to the well, and a supplement industry that already sells the claim over the counter. Rattan's alternative is health maintenance and hormesis — keep the repair systems working — not an anti-aging product. That is still medicine. It is not this claim.

Key assumptions

  • No current human intervention treats aging as such rather than a named disease or risk factor partial
  • Disease-by-disease care plus prevention is capable of most of the healthspan geroscience promises partial

Red team — the strongest counters

Disease-by-disease is the long way around — that is the geroscience claim

Stents and oncology are the wreckage strategy. Granting rapamycin in mice and then saying 'show the human hammer' is a demand that the indication exist before the trial that would justify the indication. Catch-22 is not a finding.

A supplement shop is not TAME

Sinclair's consumer stack can be a circus and metformin can still be a trial. AGAINST that uses the circus to kill the indication is the same smear it would not accept against cardiology because of a vitamin ad.

Sources

Confidence, decomposed

Logical validity●●●●○
Premise support●●●○○
Representativeness●●●●●
Source quality●●●●●

Provenance

Generated by a paired steelman agent (single model family) · red-teamed by an independent adversarial agent · sources Pass-1 spot-checked (existence and rough fit) — framing-fidelity not independently verified. Judged on merit: per the founding rule of this project, AI authorship is disclosed at site level and arguments stand or fall on their content.
AGAINST · Aging is a disease we should treat
Logically validP1

Aging is the increase in vulnerability — that is not a diagnosis

Leonard Hayflick (PLoS Genetics, 2007): the fundamental aging process is not a disease; it increases vulnerability to disease. Suresh Rattan (Aging and Disease, 2014): there are no gerontogenes with the evolved function of causing aging; considering aging a disease that happens to everybody is an oxymoron. Those are not wellness blogs. They are the biogerontology AGAINST. López-Otín's hallmarks are real experimental objects. AGAINST can grant every hallmark and still say: a list of things that go wrong with time is a description of organismal life after the evolutionary warranty, not a ICD diagnosis. Everyone who lives long enough ages; not everyone who lives long enough gets a given disease. Medicine treats the diseases. It does not owe a cure to a process whose only 'patient' is the species. Progeria is a specific laminopathy. Treating it as a warrant to medicalise typical eighty-year-olds is the analogue FOR wants and the one the genetics does not give.

Key assumptions

  • Universality (it happens to everyone who lives long enough) is a reason not to call a process a disease untestable
  • The hallmarks are better read as maintenance failure than as a single treatable pathology partial

Red team — the strongest counters

Universality did not stop us treating tooth decay or atherosclerosis

Plenty of medicine is for processes that, given time, show up in almost everyone. 'Happens to everybody' is not the clinical test AGAINST needs — 'you can be well without it' is, and the 80-year-old is not well in the way the 30-year-old is.

Increased vulnerability is what a disease process is

Hayflick's sentence can be read as FOR: the thing that raises the hazard is the target. Calling that a category error is a definition of 'disease' that geriatrics already does not use when it treats frailty.

Sources

Confidence, decomposed

Logical validity●●●●●
Premise support●●●●○
Representativeness●●●●●
Source quality●●●●●

Provenance

Generated by a paired steelman agent (single model family) · red-teamed by an independent adversarial agent · sources Pass-1 spot-checked (existence and rough fit) — framing-fidelity not independently verified. Judged on merit: per the founding rule of this project, AI authorship is disclosed at site level and arguments stand or fall on their content.
AGAINST · Aging is a disease we should treat
Logically validP1

WHO already answered: ICD is not classifying old age as a disease

WHO's ICD FAQ on 'Old age' is the official sentence: ICD is not classifying old age as a disease; the category is for quality control when a death certificate or record is vague. After protest, MG2A's title became 'Ageing associated decline in intrinsic capacity,' and XT9T's wording was pulled back from 'pathological' (Rabheru et al., Lancet Healthy Longevity, 2022). FOR's 'the codes are already in the book' is true and not the claim. The claim is that aging is a disease we should treat. A causality extension code is how you tag a stroke as ageing-related. It is not a licence to trial a drug on everyone over 65 for 'aging.' TAME exists because the FDA does not have that indication — Barzilai's public point is to convince the agency. AGAINST's point is that the agency's hesitation is the classification working as designed. Medicalising a universal process is how you make every living adult a patient. That is a regulatory and ethical choice, and WHO already declined to make it in the book FOR wants to cite.

Key assumptions

  • WHO's FAQ is a scientific classification judgment, not only a political retreat from 'old age' as a slur untestable
  • An aging indication would medicalise the well elderly in a way disease-by-disease indications do not partial

Red team — the strongest counters

The FAQ is a political retreat, not a lab result

MG2A was renamed because 'old age' sounded like a slur on a death certificate. XT9T survived. That is the biology staying in the book while the title was cleaned. WHO declined a slogan, not a mechanism.

TAME exists because the remaining fight is paperwork

Barzilai's point is the indication, which is exactly FOR's claim. Using the current absence of that indication as a finding that aging is not a medical object is how incumbency masquerades as classification science.

Sources

Confidence, decomposed

Logical validity●●●●●
Premise support●●●●●
Representativeness●●●●●
Source quality●●●●●

Provenance

Generated by a paired steelman agent (single model family) · red-teamed by an independent adversarial agent · sources Pass-1 spot-checked (existence and rough fit) — framing-fidelity not independently verified. Judged on merit: per the founding rule of this project, AI authorship is disclosed at site level and arguments stand or fall on their content.
AGAINST · Aging is a disease we should treat
Plausible, low testabilityP1

The extra years may be the disabled years — compression is a hope, not a finding

Fries's compression of morbidity is the brochure. The demographic record is mixed: some countries compressed, some lengthened the disabled tail as they cut the fatal infarct. A geroscience drug that slows every hallmark a little can, in a model, postpone death and postpone the nursing home by different amounts. Mouse lifespan is not mouse healthspan in every strain and every dose. AGAINST does not need to claim that extra life is always worse. It needs the admission that 'treat aging' does not name which years you get. Cardiology already extended the years after the first heart attack. Some of those years are good. Some are not. An indication that applies to everyone over 50, before a disease has declared itself, is how you medicalise the well and hope the tail shortens. Who gets the years is in the scope. A longevity stack that costs what consumer NAD products already cost, let alone what a monoclonal would cost, is a rich-world remaining-life purchase. Insulin became generic. Many biologics did not. The claim's distribution problem is not a footnote. It is who the extra years are for.

Key assumptions

  • Without a demonstrated compression, extra years are as likely to be frail years as healthy ones partial
  • Unequal access to a longevity indication is a reason to refuse the indication, not only to regulate price untestable

Red team — the strongest counters

The tail already expanded under disease-by-disease care

If extra frail years are the fear, the status quo is already that fear. Treating the shared cause is the attempt to compress. Refusing the attempt because the tail might grow is how you keep the tail you have.

Unequal access is a price fight, not an indication fight

Insulin, ART, and dialysis had the same distributional row. We did not leave diabetes off the list so that nobody could be a customer. Who gets the years is a delivery problem FOR can grant without losing the claim.

Sources

  • Geroscience: linking aging to chronic disease Kennedy et al., Cell 2014. Healthspan is the stated aim; it is not a demonstrated human result. Pass-1: PubMed 25417146 exists. AGAINST uses the gap between aim and evidence. P1 checked
  • TAME trial design Multi-morbidity endpoint is how FOR hopes to show compression. Pass-1: afar.org/tame-trial exists. Not a result. P1 checked

Confidence, decomposed

Logical validity●●●●○
Premise support●●●○○
Representativeness●●●●●
Source quality●●●●○

Provenance

Generated by a paired steelman agent (single model family) · red-teamed by an independent adversarial agent · sources Pass-1 spot-checked (existence and rough fit) — framing-fidelity not independently verified. Judged on merit: per the founding rule of this project, AI authorship is disclosed at site level and arguments stand or fall on their content.
AGAINST · Aging is a disease we should treat
Unfalsifiable / philosophicalP1

Natural is not the argument — 'everyone is a patient' is

AGAINST should not hide in the naturalistic fallacy. FOR is right that 'natural' is not a stop. The stop is different: disease, in clinical practice, is a condition some people have and some do not, at a given age, that you can be well without. Aging fails that test on purpose. Once it is an indication, the well 55-year-old is pre-sick, the insurance form has a new box, and the endpoint of a life is a treatment failure. That is not kindness to old people — geriatrics already owed them care. It is a change in what a human life is allowed to be without a prescription. de Grey's public FOR and Sinclair's are useful here as the voices that mean it: indefinite postponement, escape velocity, a stack you buy. The steelman AGAINST does not need them to be frauds. It needs the implication of success. A world that 'treats aging' as it treats hypertension is a world in which not treating it is non-compliance. Medicine has earned many such expansions. This one is the remaining whole of life. WHO's refusal to call old age a disease is, on this reading, the last line in front of that expansion — not a confusion about hallmarks.

Key assumptions

  • A process that includes everyone who lives long enough should not be an indication, even if it has mechanisms untestable
  • Disease-by-disease medicine can be kind to old people without this category change partial

Red team — the strongest counters

Hypertension is already 'everyone a patient' past a certain age

Rich-country guidelines medicalised the well on a number. Aging as indication is the same shape. If that is a takeover of the life course, it already happened, and we did not put blood pressure back in the forest.

Not treating is also a choice about what a life is

Leaving aging unnamed so that dying of the process is 'nature' is the value AGAINST is defending. FOR named it. The clinic-takeover fear has to beat the years of dementia that disease-by-disease care already buys.

Sources

  • WHO ICD FAQ: 'Old age' The official refusal to make old age a disease. Pass-1: already used; this arg uses it as a line about the life course, not only as a coding note. P1 checked
  • Aging Is Not a Disease (Rattan) Rattan 2014: disease-oriented anti-aging as economically and psychologically unsustainable. Pass-1: PMC 4037311 already used. P1 checked

Confidence, decomposed

Logical validity●●●●●
Premise support●●●○○
Representativeness●●●●●
Source quality●●●●●

Provenance

Generated by a paired steelman agent (single model family) · red-teamed by an independent adversarial agent · sources Pass-1 spot-checked (existence and rough fit) — framing-fidelity not independently verified. Judged on merit: per the founding rule of this project, AI authorship is disclosed at site level and arguments stand or fall on their content.
FOR · Aging is a disease we should treat
Empirical — moderateP1

Aging has mechanisms you can name — that is already a medical target

López-Otín, Blasco, Partridge, Serrano and Kroemer set out nine hallmarks in Cell (2013): genomic instability, telomere attrition, epigenetic alteration, loss of proteostasis, deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem-cell exhaustion, altered intercellular communication. The 2023 update ('An expanding universe') adds disabled macroautophagy, chronic inflammation, and dysbiosis, and restates the test: a hallmark should appear with age, worsen aging when accentuated, and offer a lever when treated. That is the same logical shape as a disease process in a textbook, minus the ICD sticker. FOR does not need the list to be final — twelve is already an expansion — and does not need Sinclair-lab press to be gospel. It needs the admission that aging is not a date on a calendar. It is a set of failures you can intervene on in mice, in cells, and, in a few cases, in drugs people already take. If we treat 'natural' as a stop, we should say so. The biology does not.

Key assumptions

  • A process that meets the three hallmark tests in animals is the right object of human medicine, not only of biogerontology partial
  • The hallmarks are causes of organismal aging rather than correlated damage partial

Red team — the strongest counters

A list of things that go wrong with time is not a diagnosis

Hayflick's point: hallmarks can be real and still be the warranty running out, not a disease some people get. Mouse levers are not a human indication. Twelve is already an expanding list — the opposite of a pinned-down pathology.

Accentuating a hallmark in a mouse is not treating aging in a person

The three-premise test is experimental rhetoric. Human aging is slower, more heterogeneous, and already wrapped in the diseases we treat. FOR's 'knobs' are still mostly animal knobs.

Sources

  • The Hallmarks of Aging López-Otín et al., Cell 153(6):1194–1217, 2013. Nine hallmarks; three-premise test. Pass-1: Cell/PubMed 23746838 exist (doi 10.1016/j.cell.2013.05.039). P1 checked
  • Hallmarks of aging: An expanding universe López-Otín et al., Cell 186(2):243–278, 2023. Twelve hallmarks. Pass-1: PubMed 36599349 exists (doi 10.1016/j.cell.2022.11.001). P1 checked

Confidence, decomposed

Logical validity●●●●●
Premise support●●●●○
Representativeness●●●●●
Source quality●●●●●

Provenance

Generated by a paired steelman agent (single model family) · red-teamed by an independent adversarial agent · sources Pass-1 spot-checked (existence and rough fit) — framing-fidelity not independently verified. Judged on merit: per the founding rule of this project, AI authorship is disclosed at site level and arguments stand or fall on their content.
FOR · Aging is a disease we should treat
Empirical — moderateP1

The diseases of old age share a cause; treating each one is the long way around

Kennedy, Berger, Brunet and colleagues (Cell, 2014) named the geroscience hypothesis in public: aging is the dominant risk factor for the chronic diseases that fill hospitals, and delaying aging is how you delay several of them at once. That is not 'be kind to old people.' It is a trial design. The TAME proposal (Targeting Aging with Metformin; Barzilai / AFAR) is the live specimen — a multi-morbidity endpoint in people 65–79, aiming at an FDA conversation about aging as an indication, using a drug already given for diabetes. TAME is not a completed result. FOR should not pretend it is. The logic does not wait for TAME: if you only pay for 'cancer' and 'heart failure,' you force every geroprotective molecule through a disease-by-disease maze, which is how a process with hallmarks stays unofficial. Rapamycin extended lifespan in genetically heterogeneous mice (Harrison et al., Nature, 2009). That is an animal fact, not a prescription. FOR's claim is the indication. Without it, the extra years we already buy with cardiology and oncology are the wreckage strategy. With it, you can try to treat the clock.

Key assumptions

  • A drug that delays several age-related diseases in humans would be delaying aging, not only those diseases partial
  • Mouse lifespan extensions (rapamycin, calorie restriction) are informative enough to justify a human indication fight partial

Red team — the strongest counters

TAME is a proposal, rapamycin is a mouse

The geroscience hypothesis is a bet. The human hammer is not on the table. Treating 'aging' without that hammer is a renaming of prevention we already do with blood pressure and statins.

Multi-morbidity as an endpoint can still be disease-by-disease in a trench coat

If metformin delays diabetes and some cancers, that is a diabetes drug with extra readouts — not proof you treated aging. The FDA conversation FOR wants may never need the word.

Sources

Confidence, decomposed

Logical validity●●●●●
Premise support●●●○○
Representativeness●●●●●
Source quality●●●●●

Provenance

Generated by a paired steelman agent (single model family) · red-teamed by an independent adversarial agent · sources Pass-1 spot-checked (existence and rough fit) — framing-fidelity not independently verified. Judged on merit: per the founding rule of this project, AI authorship is disclosed at site level and arguments stand or fall on their content.
FOR · Aging is a disease we should treat
Logically validP1

WHO already codes ageing-related processes — 'not a disease' is a FAQ

ICD-11 carries XT9T as an extension code for ageing-related causation, and MG2A, after a public row, is labelled 'Ageing associated decline in intrinsic capacity' rather than 'old age.' WHO's own FAQ is blunt: ICD is not classifying old age as a disease; the category is for quality control of vague death certificates. FOR must not invert that FAQ. The 2018–2022 episode is the opposite lesson: a title that sounded like a diagnosis was withdrawn under pressure, and the extension code's wording was softened from 'pathological' toward 'biological' (Rabheru et al., Lancet Healthy Longevity, 2022). That is politics on a classification, not a lab result. Zhavoronkov (2015) and others had already argued for using ICD-11 to make aging a target. FOR's steelman uses what survived: the system already needs a way to say 'this is ageing-related,' because clinicians already see a process. The FDA/EMA indication problem is the live bottleneck — you cannot easily trial a drug 'for aging' if aging is not an indication. Codes are not a cure. They are the paperwork that tells you the fight is about permission, not metaphysics.

Key assumptions

  • An extension code and a symptoms category are evidence that aging is a medical object, not only a billing convenience partial
  • Without an aging indication, useful geroprotective trials will not be run at scale partial

Red team — the strongest counters

The FAQ is the finding

WHO wrote that ICD is not classifying old age as a disease because people were doing exactly what FOR is doing with XT9T. An extension code is a tag on a stroke, not an indication for the well.

The indication fight is also a market

A new box on the form is how every 55-year-old becomes a customer. Regulatory permission is not a lab result. TAME's political purpose is the tell.

Sources

  • WHO ICD FAQ: 'Old age' WHO: ICD is not classifying old age as a disease; MG2A relabelled 'Ageing associated decline in intrinsic capacity.' Pass-1: who.int FAQ page exists. FOR cites it as the official limit, not as a FOR warrant. P1 checked
  • How 'old age' was withdrawn as a diagnosis from ICD-11 Rabheru et al., Lancet Healthy Longevity 2022. MG2A title change; XT9T wording. Pass-1: Lancet page exists (doi 10.1016/S2666-7568(22)00102-7). P1 checked

Confidence, decomposed

Logical validity●●●●○
Premise support●●●●○
Representativeness●●●●○
Source quality●●●●●

Provenance

Generated by a paired steelman agent (single model family) · red-teamed by an independent adversarial agent · sources Pass-1 spot-checked (existence and rough fit) — framing-fidelity not independently verified. Judged on merit: per the founding rule of this project, AI authorship is disclosed at site level and arguments stand or fall on their content.
FOR · Aging is a disease we should treat
Plausible, low testabilityP1

If the extra years can be healthy, refusing the indication is a rationing choice

James Fries's compression-of-morbidity idea is the humane version of extra life: postpone the disabled years faster than you postpone death, so the dying is shorter, not longer. Geroscience is that idea with molecular targets. FOR should not promise it. Animal interventions sometimes stretch frailty as well as life. The honest claim is that treating aging is how you even try to compress, whereas treating one cancer at a time is how you get a person who survives the infarct and lives into the dementia. Who gets the extra years is in the scope on purpose. Insulin, antiretrovirals, and dialysis were also going to be rich-world toys. Some of them became generic. Inequality is a reason to care about price and delivery. It is not a reason to keep aging off the indication list so that nobody is allowed to try. de Grey's 'longevity escape velocity' and Sinclair's consumer stack are the marketing department. They are not the claim. The claim is: aging is a medical indication we should treat. The alternative is to keep paying, at huge cost, for the diseases it causes, one at a time, and to call that nature.

Key assumptions

  • A successful aging indication would on net compress morbidity rather than extend the disabled tail partial
  • Distributional harm from unequal access is smaller than the harm of not developing the tools untestable

Red team — the strongest counters

Compression is the brochure; the tail is the record

Some countries extended disabled years as they cut fatal infarcts. A hallmark-wide drug can postpone death and the nursing home by different amounts. FOR should not cash Fries as a finding.

Insulin-became-generic is not a longevity-biologic forecast

Who gets the years is in the scope because the stack already has a shop. A monoclonal for senescence will not be metformin. Inequality is not a footnote you fix after the indication.

Sources

  • Geroscience: linking aging to chronic disease Kennedy et al., Cell 2014. Healthspan as the point of delaying aging. Pass-1: already used; this arg uses the morbidity-cluster claim, not the trial logistics. P1 checked
  • TAME trial — multi-morbidity endpoint AFAR: TAME's design is delay of several age-related diseases, not a single-cancer endpoint. Pass-1: afar.org/tame-trial already used. Design is the specimen of compression-as-trial, not a result. P1 checked

Confidence, decomposed

Logical validity●●●●○
Premise support●●○○○
Representativeness●●●●●
Source quality●●●●○

Provenance

Generated by a paired steelman agent (single model family) · red-teamed by an independent adversarial agent · sources Pass-1 spot-checked (existence and rough fit) — framing-fidelity not independently verified. Judged on merit: per the founding rule of this project, AI authorship is disclosed at site level and arguments stand or fall on their content.
FOR · Aging is a disease we should treat
Unfalsifiable / philosophicalP1

'Natural' never licensed smallpox, and it does not license untreated ageing

Bulterijs, Hull, Björk and Roy (Frontiers in Genetics, 2015) put the comparison on the table: Hutchinson–Gilford progeria is treated as a disease; when similar damage arrives at eighty it is called ageing and left off the indication list. Hayflick's reply (PLoS Genetics, 2007, and elsewhere) is the AGAINST classic: aging is not a disease, it is a process that increases vulnerability to disease, entropy in a soma that evolution stopped paying for. FOR can grant Hayflick's evolutionary story in full. Evolution's indifference is not a clinical instruction. Smallpox was natural. So is atherosclerosis. Medicine's job, on the geroscience view, is the damage, not the pedigree. Aubrey de Grey and David Sinclair are the loud public FOR voices — SENS damage-repair, NAD/sirtuin stories, the book Lifespan. They are not the evidence. They are why the public fight exists. The steelman does not need their strongest claims. It needs the weaker one that survives ridicule: if a 40-year-old and an 80-year-old differ by processes we can already name, refusing to treat those processes because they are 'natural' is a value, not a finding.

Key assumptions

  • Universality ('it happens to everyone') is not a reason to withhold treatment untestable
  • Progeria is a fair analogue for typical aging rather than a distinct genetic disease partial

Red team — the strongest counters

Smallpox was a pathogen some people got

Universality is the difference FOR wants to smash past. Progeria is a laminopathy. Typical aging is not a mutation you can analogise into an indication without remainder.

de Grey and Sinclair are not a remainder you can discard after using their fight

The public FOR is indefinite postponement and a consumer stack. A steelman that lives only in Cell reviews, with the loud voices treated as marketing, is not the claim people are being asked to buy.

Sources

Confidence, decomposed

Logical validity●●●●●
Premise support●●●○○
Representativeness●●●●●
Source quality●●●●●

Provenance

Generated by a paired steelman agent (single model family) · red-teamed by an independent adversarial agent · sources Pass-1 spot-checked (existence and rough fit) — framing-fidelity not independently verified. Judged on merit: per the founding rule of this project, AI authorship is disclosed at site level and arguments stand or fall on their content.