How the Longevity Drugs That Work in Animals Are Failing in Human Trials

Dozens of compounds extend lifespan in mice. Almost none have worked in human trials. Here is the biology behind the translation gap and the researchers who think they know how to cross it.

The model organism translation problem is one of biology's most persistent frustrations: compounds that consistently and dramatically extend lifespan in model organisms — yeast, nematodes, fruit flies, mice — routinely fail to produce comparable effects in human clinical trials. The history of longevity compound development is largely a history of mouse results that didn't translate.

Resveratrol — the polyphenol compound found in red wine that extended lifespan in yeast and worms, and became the basis for a supplement industry estimated at $300 million annually — produced no significant lifespan benefit in mice when studied rigorously at appropriate doses and under conditions that controlled for the confounders that initial studies didn't adequately address. Subsequent human trials found no longevity benefit at doses achievable through supplementation.

The biological reasons for poor translation: mice age differently from humans in ways that are more than proportional. Mice live 2-3 years and die primarily from cancer and infection. Humans live 70-90 years and die primarily from cardiovascular disease, cancer, and neurodegeneration. A longevity intervention that targets the most common cause of mouse death will not necessarily target the most common causes of human death. The biology of aging is not identical across species despite the conservation of many aging-related cellular mechanisms.

The specific compounds that longevity researchers are most cautiously optimistic about for human translation: rapamycin, whose mechanism (mTOR inhibition) is deeply conserved across species and has shown the most consistent lifespan extension results across the widest range of model organisms; senolytic drugs (dasatinib + quercetin, navitoclax) that clear senescent cells whose accumulation drives tissue dysfunction; and NAD+ precursors (NMN, NR) that address the NAD+ decline in aging cells whose consequences for mitochondrial function and DNA repair are well-mechanised.

The methodological limitation that makes human longevity research extraordinarily difficult: you cannot do a lifespan study in humans. The gold standard evidence — randomised controlled trial comparing intervention to placebo and measuring lifespan — requires 50-80 years and cannot be the primary evidence base for intervention decisions made today.