Aging clocks and faster longevity trials

Why Proving a Drug Slows Aging Waits on a Biological Clock Regulators Can Trust

This page is for learning and research. It describes scientific tools and regulatory debates and is not medical advice.

The scarcest resource in the fight against aging is not ideas for drugs. It is the decades it takes to prove that any of them work. A trial that waits for people to develop heart disease, cancer, dementia, or to die needs thousands of volunteers, many years, and tens of millions of dollars for every candidate. This page serves the futurism and abundance sides of this wiki: it explains how a trusted measure of biological aging, often called an aging clock, could turn the slow and costly search for longevity treatments into a faster, lower cost pipeline, and why that measure does not yet exist in a form regulators accept. The thesis is simple. The bottleneck for defeating aging as a research and policy fight is now validation, the unglamorous work of proving that a clock reading predicts real health, and that work deserves far more attention and funding than new clock announcements.

Why time is the real scarcity

Drug regulation is built around diseases. The United States Food and Drug Administration does not recognize aging itself as a treatment indication, a point stated plainly by researchers who support targeting aging. The designers of the Targeting Aging with Metformin trial, known as TAME, argue that without such an indication, companies will not develop a drug that has no business horizon, and insurers have no clear basis to pay for it.

TAME shows what the slow route looks like. It was planned for about 3,000 people aged 65 to 79 at about 14 centers, with a primary outcome of time to a new occurrence of cardiovascular events, cancer, dementia, or death. One commentary described the trial as needing 5 to 10 years and a budget of about 50 million dollars. A regulatory analysis noted that this composite design deliberately avoids the challenge of defining the aging process and validating measures to assess it. That is a reasonable workaround for one drug. It does not scale to hundreds of candidates.

How surrogate endpoints shorten trials

Medicine has solved this kind of problem before. The FDA explains that a surrogate endpoint is a measurement used in place of a direct clinical outcome when that outcome, such as a stroke, would take a very long time to study. Blood pressure is the classic case: many trials showed that lowering systolic blood pressure reduced strokes, so newer drugs can be tested on blood pressure in smaller and shorter trials. The FDA also reports that between 2010 and 2012 it approved 45 percent of new drugs based on a surrogate endpoint.

The agency sorts surrogates into three levels: candidate, reasonably likely, and validated. A validated surrogate needs a clear mechanism plus clinical data showing that changing the measurement predicts a specific clinical benefit. A validated aging clock would do for longevity research what blood pressure did for stroke prevention. A trial could ask whether a drug slows the clock over two years rather than waiting fifteen years to count deaths. That would shrink the price of testing each idea and let far more ideas be tested.

What an aging clock measures

The best known clocks read chemical tags called methyl groups on DNA. In 2013 a multi-tissue clock was built from about 8,000 samples drawn from 82 datasets covering 51 healthy tissues and cell types. It used 353 sites on the genome, and its readings were close to zero for embryonic and induced pluripotent stem cells, a hint that biological age might be reset. Later clocks were trained to predict health outcomes or the pace of decline rather than calendar age.

Predicting age is a different task from being a trustworthy trial endpoint. Three problems stand between the two.

Problem one: noise

A 2022 study found that technical noise produced deviations of up to 9 years between repeated measurements of the same sample for six prominent epigenetic clocks. The authors rebuilt the clocks using principal components, a statistical method that filters noise, and most repeated measurements then agreed within 1.5 years. If a measurement can swing by years on a retest, a drug effect of a few percent disappears inside the error. Reliability is the first gate, and it is solvable.

Problem two: clocks disagree

The CALERIE trial randomized 220 adults without obesity to 25 percent caloric restriction or their usual diet for two years. A later analysis of blood samples found that restriction slowed the pace of aging measured by a clock called DunedinPACE by about 2 to 3 percent, while two other clocks, PhenoAge and GrimAge, showed no significant change. The same intervention, in the same people, produced different answers depending on the clock. The authors themselves wrote that a conclusive test will require long follow-up on chronic disease and mortality.

Problem three: no proven link to outcomes

A 2023 framework from researchers working on aging biomarkers states that there are not yet any formally validated or likely surrogate endpoint biomarkers of aging. Showing that a clock responds to an intervention is only the first step, examples of failed surrogates exist, and validation is harder because there is still no consensus on which interventions improve healthspan or lifespan. This is a circular trap. To validate a clock, researchers need an intervention known to slow aging. To prove an intervention slows aging quickly, they need a validated clock.

Breaking the circle

The way out is to attach clock measurements to the slow trials that are already running or planned. If TAME or a similar outcome trial stores blood samples at the start and at regular points, researchers can later check whether people whose clocks slowed were also the people who avoided disease. One well run outcome trial with banked samples could validate, or reject, several clocks at once. That turns a single expensive trial into infrastructure for every trial that follows.

Several practical steps follow from this:

• Fund reliability work and publish test and retest numbers for every clock used in a trial.
• Bank blood samples in large outcome trials and long running cohorts so clock readings can later be matched to real events.
• Pre-register which clock will be the endpoint before a trial starts, so researchers cannot pick the clock that happened to move.
• Use the FDA process for discussing novel surrogate endpoints early, as the agency invites sponsors to do.

Why this matters for abundance

Healthy years are the most valuable thing people lose to aging, and today the supply of proven ways to protect them is tiny. A trusted clock would not itself extend anyone's life. It would make testing far less expensive, which means more candidates tested, more failures discovered early, and a faster path to treatments that work. It would also let public and nonprofit funders, not only large companies, run meaningful longevity trials, because a two year study is within reach of far more institutions than a fifteen year one.

Limits of this page

This page summarizes published research and regulatory documents. It does not recommend any test, supplement, diet, or drug. Clock results for an individual are not a diagnosis, and the reliability findings above suggest that a single reading should be interpreted with caution. Regulatory positions can change, and new validation studies may have appeared after this page was written.

Discussion questions, essay ideas, and learning related AI prompt ideas

• Explain why a measurement that predicts age well can still fail as an endpoint for a drug trial.
• Design an outcome trial that would validate or reject three competing aging clocks at the same time, and estimate what samples it would need to store.
• Compare blood pressure as a surrogate for stroke with an aging clock as a surrogate for healthy lifespan, and identify where the comparison breaks down.
• Argue whether public agencies, nonprofits, or private companies should pay for the validation work that would benefit every longevity company at once.
• Describe how the circular problem of needing a proven intervention to validate a clock could be broken by animal studies, human cohorts, or existing drugs.
• Assess the risks of approving longevity drugs on a surrogate that later turns out not to predict real health.

See also

• Dog aging trials
• Animals that barely age
• Longevity escape velocity
• Abolish aging
• Prediction markets for replication
• Adversarial collaboration
• Aging