What counts as a longevity drug?

Understanding the aging/longevity field via four distinct definitions

It’s time for the aging/longevity field to acknowledge that different people & groups within the field have different definitions of what counts as a longevity drug, or more generally an aging or longevity intervention. Here is a list of distinct definitions embraced in different contexts.

All of these have some validity and represent a better approach to dealing with age-related degeneration and diseases than the traditional “sick care” model that dominates current clinical medicine. It isn’t necessary for everyone in the field to coalesce on just 1 of these. It is important for more people to realize that others may be using a different version than they are. Too often people talk as if they don’t even realize some of the other definitions exist, and that’s nonoptimal. If you only thought about 1 or a couple of these, consider the others for a few minuets.

Four definitions

Here are 4 distinct definitions of an aging or longevity intervention:

Extends Lifespan (EL): An intervention that by itself extends lifespan (and healthspan, but not just healthspan), in normal study populations. Normal study populations for mice means normal lab mice fed normally. For humans it means modern, developed-country (e.g., US) populations, i.e. the populations most normal clinical trials are currently drawn from.

Extends Lifespan Universally (ELU): An intervention that by itself extends lifespan (and healthspan, but not just healthspan) universally in all reasonable strains & conditions. For mice universality means it would extend lifespan even of long lived strains or even of wild mice who survived predation and other environmental hazard. For humans, it means working for all normal human populations regardless of culture, geography, or historical era.

Mitigates Aging Pathology (MAP): An intervention that treats some age-related pathology that underlies several diverse age-related diseases, thereby mitigating all of these diseases. I.e., an intervention that embodies the geroscience hypothesis. The TAME trial’s multimorbidity endpoint is essentially a special case of this. Extending healthspan (without necessarily extending lifespan) fits best into this definition, though see below.

Indefinite Lifespan Necessity (ILN): An intervention that successfully treats an age-related pathology that must eventually be treated in order to fully eliminate adult biological aging and age-related diseases or equivalently in order to achieve indefinite lifespans. Ie, an intervention that fixes something that would eventually cause age-related pathology & death even if all other aspects of aging were fully cured.

These are not necessarily exhaustive but they cover several important views. Implicitly these partly define different approaches to the field itself, and discussion of their specific differences and how notable things in the field relate to them is a useful lens by which to view and understand some of the different paradigms and specific major efforts within the field.

Necessary vs. sufficient

Definition Extends Lifespan (EL) vs. definition Indefinite Lifespan Necessity (ILN) is in essence the difference between sufficient vs. necessary conditions. EL essentially means sufficient to move the needle on longevity at least a bit from its current state, whereas ILN means necessary to move the needle the last bit to get to the largest possible effect size (aging fully cured) even if we’re almost all the way there. Note: necessary here refers to the job not the intervention, so it means does a job that is necessary not that the intervention is the only possible intervention capable of doing that job.

Something that satisfies EL may not satisfy ILN, and vice versa. Eg, something that slows all aspects of aging by 10% but fully stops or reverses none of them will satisfy EL but not ILN. in contrast, something that fully fixes an aging sub-pathology that underlies many age-related diseases but doesn’t extend lifespans by itself (because other sub-pathologies still cause decline and death just about as quickly) will satisfy ILN but not EL. This distinction is not commonly discussed in the field. In fact, definition 4 is not commonly discussed explicitly.

Examples of interventions that satisfy EL abound, at least for mice. Calorie Restriction (CR) and rapamycin are classic examples. These probably do not satisfy ILN. The classic example of an intervention satisfying ILN is ways to break down or eliminate from the body forms of lipofuscin that are undegradable by any enzymes coded in human DNA and also not naturally exported from the body, as described by Aubrey de Grey in the 2000s. This kind of lipofuscin may build up so slowly that it doesn’t make notable contributions to age-related health decline over the course of current normal lifespans, but some solution will eventually be needed if lifespans grow long enough for its accumulation to become toxic.

Universality and GLP-1 drugs

Definition Extends Lifespan Universally (ELU) is just like definition Extends Lifespan (EL) but imposes stricter universality in terms of what populations would benefit.

Obviously, the issue of whether GLP-1 drugs count as longevity drugs has been discussed for a while in the field, culminating in the recent Nature Biotechnology editorial “​​Are GLP-1s the first longevity drugs?” Whether GLP-1 drugs are longevity drugs depends partly on which of the above definitions you test them against. GLP-1 drugs satisfying EL looks likely, whereas satisfying ELU is possible but currently looks less likely or at least there is no current data testing them in lean populations that I’m aware of. This means they might do nothing useful to cohorts with very low rates of overweight/obesity. For example, Vietnam has ~2% obesity rate vs. ~42% in US. The average person in the US in the mid-20th-century had BMI below 22.5. Possibly GLP-1 drugs could have efficacy in cohorts like these, but we don’t know yet. Even in mice we don’t know whether lean mice would show any benefits.

For other interventions in general, extending average lifespans of the population that study subjects are normally drawn from is good enough for EL, but interventions that only fix age-accelerating pathologies that aren’t universal features of human aging will not satisfy ELU.

Healthspan vs. lifespan

Definition Mitigates Aging Pathology (MAP) means just that, mitigating one or more of the pathological aspects of aging biology. Interventions that satisfy MAP will improve healthspan but won’t necessarily by themselves extend lifespan and satisfy definition EL. Also, not everything that improves health necessarily qualifies as MAP. It must do so by treating some aspect of aging pathology. Thus, this definition is phrased as mitigating aging pathology rather than just extending healthspan to avoid pulling in the entire field of health improvement or wellness, much of which is not traditionally considered part of the aging/longevity field.

Healthspan or healthy lifespan are also vague terms that makes objective measurement hard to operationalize. In essence they rely on a definition of (good) health or if that is taken to be freedom from disease or significant pathology then on a definition of these latter concepts and in fact not just on a definition but implicitly on a threshold dividing good health from bad or significant disease/pathology from normal variation, but in either case the underlying biological state really exists on a continuum.

Nonetheless, one can impose thresholds or use the existing thresholds of clinical diagnoses of diseases per current medical practice today. This is what the TAME trial’s multimorbidity endpoint does, using age-related chronic diseases.

Hallmarks, SENS, ITP, X-Prize, etc.

Many major efforts or paradigms within the field implicitly or explicitly embrace a specific one of these definitions.

The Hallmarks of aging were defined based on criteria that implicitly embrace definition MAP: Criteria 1 to qualify as a hallmark said the phenomenon has to manifest during normal aging (i.e., be an aging pathology) and criteria 3 said that ameliorating the phenomenon had to slow aging and thereby extend healthspan.

In contrast, the earlier SENS framework (paper full text) implicitly embraces definition ILN: Each of the 7 damage areas were meant to be aging pathologies that must be remedied to achieve full elimination of aging and thereby lifespans not limited by aging.

The geroscience hypothesis most closely matches to definition MAP. Treating fundamental biological mechanisms (aging pathologies) can mitigate/delay/treat multiple age-related chronic diseases or age-related decline. This will extend healthspan and may or may not extend lifespan, with the presumption that treating all such biological mechanisms will eventually extend lifespan too.

The NIH’s Interventions Testing Program (ITP) inherently embraces definition EL. Extending lifespan is the main endpoint. A criticism of mouse lifespan work is that they die of cancer so predominantly that things that extend their lifespans may be only mitigating cancer specifically rather than a broader aging subpathology, so the worry is that some things that do well in ITP may not satisfy MAP, though many (like CR and rapamycin) if not most probably do.

Similarly, platforms that screen large numbers of compounds for lifespan extensions, as used by many academic projects and biotech companies, implicitly embrace definition EL.

The Longevity Biotechnology Association defined a longevity biotechnology company using a 4-pillar criteria: mission to combat age-related diseases and extend healthspan, method to target aging mechanisms (subpathologies), regulatory route via clinical trials for 1 then multiple age-related diseases, and finally multimorbidity trials (same endpoint as TAME). These criteria are essentially the same as definition MAP.

The X-Prize Healthspan, like the TAME trial and the LBA aging biotech company definition, does not explicitly test lifespan. It’s multi-pronged functional reversal of decline also best satisfies definition MAP.

The (whole-body) replacement approach to aging, as championed by Jean Hebert, tries to sidestep the issue of fixing each aging subpathology by simply replacing everything with new, young versions, implicitly satisfying definition ILN.

Longevity.Technology’s 10 levels of longevity framework has somewhat different goals and is somewhat orthogonal to the definitions of what counts as an aging or longevity drug here, so mapping between them is imperfect. Roughly, their level 5 (aging disease management, i.e. current mainstream healthcare) is out of scope here—it doesn’t satisfy any of the 4 definitions. Their levels 1-4 (lifestyle, consumer diagnostics, supplements, longevity clinics) are either also out of scope as just wellness (e.g., some consumer diagnostics) or match mostly to definition MAP here. Level 10 (body/mind preservation) is out of scope. Levels 6-9 are ambiguous between the definitions. Level 6 (aging disease prevention) and 7 (aging disease reversal) arguably match best to MAP, but level 7 could match to ILN. Level 8 (systemic age reversal) is could match to MAP or any of the others depending on how comprehensive the age reversal. Level 9 (organ/-ism preservation: extending human life beyond natural limits) moves more into the other definitions, EL, ELU, or ILN without necessarily being specific to only 1 of the 3.

On the importance of not ignoring the other definitions

Too many conversations, claims, or written works in the field ignore some of these 4 definitions or treat just one of them as if it is the organizing principle of the whole field. Efforts associated with all of these share a common belief that treating aging directly is better than the siloed one-disease-at-a-time model of medicine and science research funding and that treating aging preventatively is better than reactive sick-care that starts only after clinical stage age-related-disease diagnoses. Thus, advocates of each of these definitions should primarily be allies against the status quo as well as against aging, despite some differences of opinions within the field.

It is inappropriate to think of only things that extend lifespan as being “in” the aging or longevity field. It is inappropriate to say that the ITP should be the only yardstick by which progress in the field is measured, and inappropriate to lose all faith in an intervention as relevant to aging if it happens to fail in the ITP. It is inappropriate to declare GLP-1 drugs to be longevity drugs for everyone. It is inappropriate to claim that no FDA approved interventions exist for aging or are coming soon since several that satisfy definition MAP are now FDA approved and more are in late-stage trials. It is inappropriate to ignore the idea of potentially eventually fully fixing every aspect of aging (as the Hallmarks paper arguably effectively did by ignoring all prior published work on SENS).

Lastly, it is probably inappropriate, and certainly impractical, to try to get everyone in the field to agree on fundamental definitions, let alone to agree on branding/marketing angles for the field (as much as such exercises may still be valuable and their motivations are understandable, and the dedication to the field of the people leading them is beyond question). People outside the field are often drawn more strongly to some of these definitions than others, and in some cases are even repulsed by some of them but sometimes others outside the field are specifically drawn most strongly (or even only) to the very ones that repulse some other people.

Regardless, the field can and will continue to make progress scientifically, clinically, and in terms of public awareness and support.

How these definitions influence my work

Definition MAP is essentially the definition I use for AgingBiotech.info, for deciding whether something is in-scope for the aging/longevity field as a whole vs. out-of-scope (eg, just health broadly, or too disease-specific). It is essentially the definition I used throughout 2025 when discussing and giving talks on FDA approvals of aging interventions and those in phase 3 clinical trials.

Definition ILN is what I strive to invest in with my investing activities.