Longevity Research Limitations and Safety

Why this page exists

Longevity headlines often jump from "extended lifespan in mice" to "may help you live longer" in one step. This page explains the gaps between those claims and proof, and the safety questions to ask. Labels used: established, promising but limited, preclinical (cells/animals only), insufficient evidence, evidence of harm, marketing claim. Education only; it does not recommend products or tell you what to take.

Limits of the research

Animal results do not automatically carry over. The National Institute on Aging Interventions Testing Program has found agents that raise median lifespan in mice, but also many tested agents with no effect, including metformin alone, resveratrol, curcumin and fisetin in its main results. [1] Even the winners often work in one sex only: a 2025 review of the program found most compounds that extended lifespan worked mainly or only in male mice. [2] Whether similar drugs slow aging in people is unproven; a 2025 JAMA review said the health benefits of reducing senescent cells in humans remain unclear. [3] In cancer research, a review found that the average rate of successful translation from animal models to clinical trials was under 8%. [4] That figure is from cancer, not aging, so it shows the general problem rather than a measured rate for longevity drugs. See The Interventions Testing Program.

Lifespan trials in people are hard. Measuring who lives longer takes decades, so studies use shortcuts. The CALERIE analysis of aging clocks said a conclusive test needs long follow-up on chronic disease and death. [5]

Surrogate endpoints and biomarkers. A surrogate is a measure used instead of the outcome that matters. A 2024 review said no consensus exists on how to validate biomarkers of aging before they are used in the clinic. [6] Surveyed experts named six barriers to translating them, including the need for markers that work at the level of individuals. [7] A change in a blood marker, a DNA methylation score or a tracker number is not the same as living longer or staying well. See Surrogate Endpoints and Aging Biomarkers in Trials and Biomarkers and Biological Age Tests.

Small, short trials. The PEARL rapamycin trial lasted 48 weeks and its main outcome did not change significantly. [8] A review of NAD+ precursors said small sample sizes limit interpretation. [9]

Observational studies. Large cohort studies on physical activity found links to lower risk of death, but they show association, not cause. [10]

A model of a trial that checks real outcomes. ASPREE followed 19,114 healthy older adults and measured death, dementia and persistent disability together. Low-dose aspirin did not prolong disability-free survival and caused more major bleeding. [11] Hopeful ideas can fail when tested this way.

Longevity clinic and product claims

We looked for systematic reviews testing the health outcomes of longevity clinic packages (testing panels plus supplements or prescriptions) and found none that we could read. Proper evidence not found. Treat such claims as marketing claims until supported by trials in people. A 2024 review described anti-aging medicines as agents in ongoing trials against age-related diseases. [12] See Longevity Clinics: What They Offer.

Safety

• Supplements are not risk-free. In a meta-analysis of 68 trials, antioxidant supplements did not reduce death overall, and beta carotene, vitamin A and vitamin E may increase mortality. [13]
• Prescription medicines have real risks. MedlinePlus warns that sirolimus can lower resistance to infection and may raise cancer risk, and that metformin can rarely cause lactic acidosis. [14][15] See Experimental Longevity Compounds.
• Tests have limits. The FDA says direct-to-consumer tests for general wellness are generally not reviewed before they are offered, and have varying levels of evidence. [16]
• Check with a professional. Pregnancy, medicines, kidney or liver disease and older age all change risk. See older adults, medication interactions and talking with a health professional.

A checklist for claims

1. Was it tested in people, or only in cells or animals?
2. Did it measure health outcomes or only a lab number?
3. How many people, for how long, and was there a placebo group?
4. Who pays for the claim, and is there something to sell?
5. What are the risks, and who should not use it?

For what does have human support, see Longevity Interventions: What Human Evidence Exists and How We Label Evidence.