Can We Slow Human Aging?

Curing any single disease of old age adds surprisingly little to life expectancy, because the others are waiting. Eliminating all cancer, for example, would add only a few years on average. Slowing the underlying process is a different proposition: it would push back heart disease, cancer, dementia and frailty together. That is the promise of geroscience, and why a modest effect on aging could be worth more than a large effect on one disease.

To show with reasonable confidence that a drug cuts deaths by 10%, a trial needs roughly 2,800 deaths. Healthy 65-year-olds die at around 2% a year, so collecting those deaths takes about 30,000 people followed for five years (Derived). The TAME trial of metformin tried to solve this by counting a combination of age-related diseases rather than death alone, bringing its planned size down to about 3,000 people aged 65 to 79. Even that design has struggled for years to raise full funding. A sufficiently validated surrogate could allow shorter trials, but their size would depend on marker variability and treatment effect; hundreds of participants is a possibility, not a result of the survival model. That is why the measurement problem sits at the centre of the field.

A younger blood result is one rung on the evidence ladder

The strongest animal evidence comes from interventions tested in genetically diverse mice at multiple sites. Even there, effects depend on sex and dose: acarbose’s large male benefit is much smaller in females. That variation is part of the result, not statistical clutter. Moving from mice to people adds different genetics, environments, disease histories and decades of potential drug exposure. A plausible mechanism and longer mouse lifespan establish a reason to test an intervention; they do not establish a safe preventive treatment for humans. NIA’s Interventions Testing Program; NIA’s acarbose findings

CALERIE’s calorie-restriction intervention changed DunedinPACE by about 2–3%, while the study found no significant effect on the PhenoAge or GrimAge estimates. The analysis used blood DNA from 197 participants in a parent trial that randomized 220 adults. The distinction matters because clocks measure different aspects of aging. The often-quoted 10–15% mortality implication comes from associations in other cohorts, not deaths prevented in CALERIE. A valid surrogate would need to show that treatment-induced changes predict improvements in outcomes, rather than merely that healthier people tend to have better readings. CALERIE biomarker analysis, Nature Aging, 2023

Human lifespan extension has not been demonstrated by these results. A biomarker can help select candidates before it is sufficiently validated to replace a clinical endpoint.

Why the dog trial matters

Dogs live in our homes, eat similar diets and develop many of the same diseases, but their lives run several times faster. Loyal’s STAY trial is following about 1,300 senior dogs across some 70 veterinary clinics. Because older large dogs die at high annual rates, a lifespan trial is feasible with hundreds of animals rather than tens of thousands of people. Loyal reports FDA acceptance of its reasonable-expectation-of-effectiveness and target-animal-safety application sections; these milestones are not a completed efficacy trial or a final approval. If the dogs that take it live measurably longer and healthier lives, it would be the first drug approved to extend lifespan in any species, and a powerful argument for human trials.

Smaller trials require a better endpoint, a larger effect, or a higher-risk population

The calculator’s mortality example needs about 2,800 deaths and 31,000 participants to detect a 10% hazard reduction over five years, under its simplified assumptions. Counting several age-related diseases raises the event rate and can shrink the trial. Recruiting frailer volunteers does the same. Neither shortcut is free: composite components can differ in importance and treatment response, and results in high-risk patients may not generalize to healthy adults. A biomarker-based trial has separate arithmetic, based on marker variability and treatment effect; the survival model cannot calculate its size.

Dogs provide a faster real-world test of a longevity intervention, but regulatory progress and lifespan proof are separate. Loyal reports FDA acceptance of the reasonable-expectation-of-effectiveness and target-animal-safety sections for LOY-002, and full enrollment of its STAY efficacy trial. FDA veterinary conditional approval permits marketing before the full effectiveness standard has been demonstrated, subject to other requirements. Even a conditional approval would therefore leave an important question for the randomized trial: do treated dogs live longer and remain healthier? A positive canine result would strengthen the case for human studies, without determining their effect size or safety. Loyal’s February 2026 status report; FDA’s explanation of conditional approval

Participant counts are Derived. Company application milestones are reported status; no completed STAY efficacy result is asserted here. Healthspan needs functional and disease outcomes as well as survival.

Who is building what

Drug developers, trials and the funders and consortia building the field’s evidence base. Search the record, or filter by role.

Changes in biological-age clocks are early signals, not outcomes. A therapy has slowed aging only when randomised trials show fewer age-related diseases, better function or longer life.

Four numbers to watch

First, the result of the STAY dog trial and whether LOY-002 gains conditional approval. Second, whether any biomarker of aging is accepted as a trial endpoint by a major regulator. Third, the number of randomised human trials using composite age-related disease endpoints. Fourth, the gap between lifespan and healthy lifespan, the outcome the whole field exists to shrink.