Introduction

The question

Medicine has learned to count what it treats. A drug for high blood pressure is judged by blood pressure and, in the end, by strokes and deaths; a cancer drug by tumours and survival. Each has a disease that can be named, a measure that can be read within a trial, and an outcome that people care about. Ageing fits none of these slots easily. It is not, in the usual categories, a disease. Its outcomes, the illnesses and frailty of late life and death itself, arrive decades after anything a trial might do in middle age. And the measures proposed to read it within a trial, from blood chemistry to patterns of chemical marks on DNA, are new and disputed.

Yet a growing body of research argues that ageing is the thing to aim at. If the same biology drives many late-life diseases, then slowing it might delay them together, and a treatment aimed at one disease at a time would miss most of what is at stake. The research programme built on this idea is called geroscience. Whether its promise holds in people is not this paper's question, and nothing here says that it does. The paper asks the question Meridian asks of any decision: who must decide, by when, and what would make them decide again? Put in full, the question is this. When should ageing itself count as a target for a trial or a regulator, who decides what counts as evidence that it has been slowed (biomarkers, clocks of ageing, healthspan), by when, and what would make them revisit it?

The question has a subject, an object, a time and a condition, and each can fail. A measure may be adopted by people with no standing to adopt it; a measure may be chosen because it moves within a trial rather than because it tracks what people care about; a decision may be taken with no date by which anyone must look at it again. The paper reads nine public texts as partial answers to that question. It reports where each places the decision and what each leaves open. It does not take a side on any intervention, method or policy. Several of the texts report experiments in animals or trials in people; they are read here for how they frame evidence, and their findings are given as their authors state them, with the limits their authors state. Nothing in this paper is medical, clinical, nutritional, lifestyle, regulatory or investment advice.

The debate

The first debate is whether ageing is a disease at all. David Gems, The aging-disease false dichotomy: understanding senescence as pathology (2015), argues that it is, or close to it, and that the opposite view does harm. He sums up the error in his abstract:

The same essay turns on his own field in the next breath. If ageing is pathology, it does not follow that there is one process to treat, and Gems calls the hope for one a likely error:

The two sentences together set the problem this paper follows. If ageing is a disease, it can be a target; if it is not one process, it is not obvious what a trial against it would measure. The review by Carlos López-Otín and others, Hallmarks of aging: An expanding universe (2023), gives one answer from the biology: as its abstract describes them, it lists twelve hallmarks of ageing and defines a hallmark partly by the opportunity to slow, stop or reverse ageing by intervening on it. A target, on that definition, is something one can intervene on. Brian K. Kennedy and others, Geroscience: Linking Aging to Chronic Disease (2014), argue in their short commentary that because delays of ageing are possible in mammals and ageing is the greatest risk factor for most chronic diseases, research aimed at extending human healthspan should expand.

The second debate is over the goal: a longer life, a healthier one, or both. David A. Barzilai, Healthy life extension: Geroscience's north star (2026), writing as a longevity physician, refuses the choice and asks that the field be judged by health-adjusted survival. He also reports, with approval, the demand of Mikhail Blagosklonny, whom he honours, that cuts against much of the literature this paper reads:

The third debate is about value. Andrew J. Scott and others, The economic value of targeting aging (2021), place a money value on longer and healthier life with a standard economic method and compare what people would be willing to pay for each. Their abstract states the result:

The fourth debate concerns measurement, and it is the one on which the others depend. Steve Horvath, DNA methylation age of human tissues and cell types (2013), built a predictor of age from chemical marks (methyl groups) at 353 sites on DNA that works across most tissues; such predictors came to be called epigenetic clocks. Daniel W Belsky and others, Quantification of the pace of biological aging in humans through a blood test, the DunedinPoAm DNA methylation algorithm (2020), argue for a measure of the rate of ageing rather than a clock, and build one. Mahdi Moqri and others, Biomarkers of aging for the identification and evaluation of longevity interventions (2023), say in their abstract that the lack of standards and of consensus on what makes a biomarker of ageing reliable hinders its clinical validation, and they propose a framework of terms, classes and use cases.

The fifth debate is about trials. Jamie Justice and others, Frameworks for Proof-of-Concept Clinical Trials of Interventions That Target Fundamental Aging Processes (2016), report a meeting that proposed three frameworks for early trials (chronic diseases of age, geriatric syndromes, and resilience to stressors) and named, among its goals, further dialogue with regulators about which indications such a therapy could be registered for. John C. Newman and others, Strategies and Challenges in Clinical Trials Targeting Human Aging (2016), summarise a related meeting on study populations, interventions and outcomes, and how it informed the proposed TAME study. In animals, the long-running answer to the question of who chooses what gets tested is described by Richard A. Miller and others, An aging Interventions Testing Program: study design and interim report (2007): a programme of the National Institute on Aging that tests candidate agents in genetically varied mice at three sites and solicits candidates from the research community each year.

The gap and the contribution

These debates are usually conducted as arguments about what is true: whether ageing is a disease, whether a clock measures ageing, whether an intervention slows it. Each is a real question, and none is settled in the texts. This paper sets them aside and asks a prior question about the decision itself. Before anyone can say that ageing has been slowed, someone must have decided what would count as showing it, and that decision has a holder, a timetable and conditions for review, whether or not anyone wrote them down.

The contribution is to read the nine texts for those three things and to set the answers side by side. The paper finds that the texts are generous about the target and careful, sometimes severe, about the evidence; that the decision about evidence is mostly left to the research community, while one federal text addresses the regulator directly and sets a period; and that the conditions for looking again are unusually explicit in the measurement papers, which write them as the steps a marker must pass before it can stand in for the outcomes people care about. Chapter 1 gives the method and the findings; the conclusion discusses them and states the paper's limits. The appendix quotes further passages of every text read.