How We Label Evidence in Longevity Research
The plain-language evidence labels used across the Longevity and Healthy Aging library, how to tell animal, biomarker and human outcome findings apart, and how to read a longevity claim.
Content last updated . This page has not been medically reviewed.
Why labels are needed
Longevity is a research field, not a treatment category. Many headlines describe a result in cells or mice as if it were a result in people. To keep this clear, every page in this library uses the same labels. They describe how strong the evidence is for a specific claim in a specific group, not whether something is "good" or "bad". Nothing here is medical advice, and aging itself is not an approved medical indication for treatment.The labels
| Label | What it means here |
|---|---|
| Approved medicine | A regulator has approved it for a named condition (we name the regulator and the indication). This is never the same as approval for aging. |
| Established | Consistent, good-quality human evidence, often from several randomized trials or strong systematic reviews. |
| Promising but limited | Some human evidence, but small, short, a single group, or about a marker rather than health. |
| Mixed | Studies disagree, or results depend on population, dose or design. |
| Insufficient evidence | Too few or too weak human studies to say. |
| Preclinical (cells/animals only) | Seen in cells, worms, mice or other animals. Not shown in people. |
| No good evidence of benefit | Looked for in good studies and not found. |
| Evidence of harm | Studies show risk or adverse effects. |
| Marketing claim | A statement made to sell a product or service with no supporting study we could read. |
| Proper evidence not found | We looked and found nothing usable. We then say what we checked. |
Three ladders to climb
1. The species ladder. Cells, worms, mice, monkeys, then people. A BMJ systematic review compared animal experiments with human trials for six treatments. Corticosteroids showed no benefit in head injury trials but did in animal models, and tirilazad improved outcomes in animal stroke models yet was linked to worse outcomes in patients. [1] The authors said discordance may come from bias or from models that do not mimic human disease. [1] For aging, see Aging in Mice, Worms and Humans.2. The outcome ladder. A lifespan result, a clinical event (heart attack, fracture) and a biomarker (a blood level, a clock score) are different things. In the CALERIE trial, two years of 25% calorie restriction improved blood pressure, cholesterol and C-reactive protein in healthy adults. [2] Those are risk-factor changes, not proof of longer life. We call a biomarker change a biomarker change. See Healthspan and Lifespan.
3. The design ladder. Randomized trials start as higher-certainty evidence and observational studies as lower, in the GRADE system. GRADE uses four categories (high, moderate, low, very low) applied to a body of evidence, and certainty can be reduced by imprecision, inconsistency, indirectness or publication bias. [3] Observational association is not cause. A large CRP meta-analysis, for example, found associations with disease and death that depended considerably on other risk factors. [4]
Other checks we apply
- Sex, dose and start age. In the NIA Interventions Testing Program, most compounds that extended mouse lifespan worked mainly or only in male mice, and dose and start age affected results. [5]
- Preprints. A preprint is a paper posted before peer review. We say so when we cite one, as in Centenarians and Blue Zones.
- Negative and mixed findings. We include them. Examples are the monkey calorie-restriction studies that disagreed on survival [6], and the trial results discussed in Inflammaging.
- Funding. We note it when stated. The CANTOS trial, for example, was funded by Novartis. [7]
- Regulatory status. Only as read from the regulator's own page, with the date.
How to read a longevity claim
- Who or what was studied: cells, animals or people?
- What exactly was measured: lifespan, a clinical outcome or a biomarker?
- How big and how long was the study, and compared with what?
- Was it repeated by others?
- Is someone selling something based on it?
For more, see Reading Longevity Studies and Longevity Research Limitations and Safety.
Frequently asked questions
Does "promising" mean it works?
No. It means some human evidence exists but it is small, short, narrow or about a marker, so benefit is not shown.
Why do you label animal findings separately?
Because animal results often do not match human results. A BMJ review found cases where animal and human trial effects disagreed. [1]
What is GRADE?
A system that rates certainty of evidence as high, moderate, low or very low, applied to a body of evidence rather than one study. [3]
Is a biomarker improvement a health benefit?
Not by itself. A biomarker such as C-reactive protein can improve without any proof that people live longer or healthier lives. [2]
What does "Proper evidence not found" mean?
We searched for sources and found nothing usable. We then state where we looked.
More details: evidence, limits and cautions
The details behind this page: cautions, limits of the evidence and where sources disagree. Open any note to read it.
Safety note
References
- Perel P et al. Comparison of treatment effects between animal experiments and clinical trials: systematic review. BMJ 2007. PubMed 17175568. Read 2026-10-22.
- Kraus WE et al. 2 years of calorie restriction and cardiometabolic risk (CALERIE): exploratory outcomes of a multicentre, phase 2, randomised controlled trial. Lancet Diabetes Endocrinol 2019. PubMed 31303390. Read 2026-10-22.
- Balshem H et al. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol 2011. PubMed 21208779. Read 2026-10-22.
- Emerging Risk Factors Collaboration. C-reactive protein concentration and risk of coronary heart disease, stroke, and mortality: an individual participant meta-analysis. Lancet 2010. PubMed 20031199. Read 2026-10-22.
- Sex as a major determinant of pro-longevity drug efficacy: a review of two decades of the NIA Interventions Testing Program. J Gerontol A Biol Sci Med Sci 2025. PubMed 40717358. Read 2026-10-22.
- Mattison JA et al. Caloric restriction improves health and survival of rhesus monkeys. Nat Commun 2017. PubMed 28094793. Read 2026-10-22.
- Ridker PM et al. Antiinflammatory therapy with canakinumab for atherosclerotic disease (CANTOS). N Engl J Med 2017. PubMed 28845751. Read 2026-10-22.