ArticleHealthy longevity

Mouse studies and longevity: why they are not enough for people

Published 11 min read12 sources

AI-generated content. How we make it

In brief

  • Studies in cells and animals are useful for finding leads and understanding mechanisms, but they answer a different question: what happens in that species and under those conditions. That something extends a mouse's life doesn't prove that it extends a person's.
  • The step from animals to people often fails: in a large review of disease therapies tested in animals, among those with at least ten years of research behind them, half went on to be studied in people and only 5% were eventually approved.1
  • Even among mice, results vary with sex, strain and laboratory,57 and in monkeys, calorie restriction gave different answers in two parallel studies.10
  • Many headlines don't say that the study was done in mice.12 Before drawing conclusions, ask who it was done in, what was measured, what it was compared with and how much changed.

Headlines like “a drug extends life by 20%” or “a diet slows aging” are often about results that, read carefully, come from mice, worms or cells in a dish. These studies aren't useless, but they don't answer the question the reader is asking: will this work for me?

This article explains what each type of study can show, why the jump from animals to people fails so often, and what questions to ask of a headline. It's part of our guide to healthy longevity, which sums up in a diagram what each level of research can tell us; here we go into it with examples.

What each type of study can show

Each type of study answers a different question. A finding in cells shows a possible mechanism; one in animals, what happens in that species and under those conditions; an observational study in people, an association that may have other explanations; and a randomized trial, which assigns at random who receives an intervention, is the most direct way to see whether something causes an effect. None of these rungs replaces the next one.

The systems used to rate evidence reason in a similar way. In the GRADE system, randomized trials start with high confidence and observational studies with low confidence; confidence goes down, among other reasons, when the evidence is indirect, and an observational study can gain confidence if the effect is very large.3 GRADE doesn't talk about animals, but it follows from its logic that a study in mice is, for people, indirect evidence.

A practical consequence, which follows from the timescales: a trial that set out to measure how long people live would have to last decades. That's why measures that change sooner, such as “biological age” tests, attract so much interest, and why it's worth asking what they really reflect.

How many animal results make it to people

A review of reviews published in 2024 pooled 122 systematic reviews covering 367 therapies tested in animals for 54 human diseases. Among those with at least ten years since their first animal study, half made it to some study in people, 40% to a randomized trial and only 5% were approved.1

Figure 1. Of the therapies tested in animals, only one in twenty was eventually approved

281 therapies for human diseases, not for extending life, with at least ten years since their first animal study, gathered from systematic reviews published up to August 2023. Approval meant approval by the US Food and Drug Administration (FDA) or a recommendation in UK or Swiss guidelines, and approval rates ranged from 0% to 20% depending on the type of disease.1 The authors caution that reviews of this step may be more common in fields where some therapy has already worked, which would push the figures up.

An earlier, smaller study reached a similar conclusion. Of 76 highly cited animal studies, published between 1980 and 2000 in seven of the most influential scientific journals, 28 (37%) were confirmed in randomized trials in people, 14 (18%) were contradicted and 34 (45%) remained untested; eight of the confirmed interventions were eventually approved. And these were the most visible studies, which according to the authors were more likely than the rest to make it to people.2

Why is so much lost along the way? The reviews point to several reasons: the animals are usually young and healthy, unlike the people a treatment is meant for; many animal studies don't randomize the groups or assess the results blind; and the effect tends to shrink in the move from animals to people and, in at least one of those reviews, also as studies become more rigorous.1 Of the 76 highly cited studies, few assigned the animals at random or assessed the results without knowing which group each animal belonged to.2

Nor should we exaggerate in the opposite direction. In 62 well-studied therapies, the 2024 review found positive results in 79% of animal studies, around 60% of studies in people and 50% of randomized trials.1 What works in animals often works in the first studies in people; the funnel narrows mostly at the more rigorous trials and at approval.

Extending a mouse's life: what has been seen and what it doesn't tell you about yourself

Short-lived organisms allow something impossible in people: measuring an entire lifespan in a short time with everything controlled. The US National Institute on Aging's Interventions Testing Program (ITP) is designed to make those results robust: it tests each substance at the same time in three laboratories, with genetically diverse mice,4 and aims to publish all results, positive or negative.6

In 2009, that program reported that rapamycin, added to food from 600 days of age, extended the lives of male and female mice: the age by which 90% of the animals had died increased by 14% in females and 9% in males. In that study, half of the untreated females had died by around 881–895 days, a little under two and a half years, and the authors compared a 600-day-old mouse to a person of about 60, in proportion to each species' maximum lifespan.4 This is a result in mice: none of the sources in this article shows that rapamycin extends human life.

Even within that program, the results aren't uniform:

  • Sex. Acarbose increased median lifespan by 22% in male mice and by 5% in females.5
  • Laboratory. Another compound, 17-α-estradiol, extended the lives of males by 28% in one of the three laboratories and by 3% in the other two, a difference that wasn't significant at those two sites. The authors don't know why untreated males live longer at one center than at another.5
  • Negative results. Five substances given from four months of age — green tea extract, curcumin, oxaloacetic acid, medium-chain triglyceride oil and resveratrol — didn't significantly change the mice's lifespan at the doses tested.6
  • Strains. Outside that program, most longevity studies use strains with little genetic diversity: almost all studies of the worm C. elegans have used a single strain, and the effect of calorie restriction in mice depends on the strain.7

And what works in an animal may not carry over to people even when the mechanism looks similar. An example given in a perspective article: knocking out the growth hormone receptor extends the lives of mice by more than 40%, whereas, according to that article, people with a deficiency of that receptor don't live longer, although they seem better protected against cancer.7

Calorie restriction in monkeys: two studies, two answers

Rhesus macaques live for decades and age in a way similar to people. That's why, in the late 1980s, two studies began testing calorie restriction in them — eating less without becoming malnourished: one at the University of Wisconsin and another at the US National Institute on Aging (NIA).10

In 2009, the Wisconsin study reported that the restricted monkeys had fewer deaths from age-related causes and less diabetes, cancer and cardiovascular disease.8 In 2012, the NIA study reported that restriction hadn't improved survival, although both studies had documented health benefits.9 In 2017, the two teams compared their data and pointed to design differences that could explain the discrepancy.10

Differences between the two calorie restriction studies in rhesus macaques, according to their joint analysis10
AspectWisconsinNIA
Monkeys76121
Age at the start7 to 15 years, already adults1 to 23 years, in age groups
Control group's foodRoughly what they ate when allowed to eat freelyA fixed ration, calculated from tables based on age and weight, without being allowed to eat freely
DietSemi-purified; 45% of the carbohydrates were sucroseMade from natural ingredients; less than 7% of the carbohydrates were sucrose
Survival with restrictionBetter: the control monkeys had almost twice the death rateNo significant differences from the controls

The authors' joint interpretation is that restriction improved health in both studies, and survival when it began in adulthood. At the NIA, the monkeys that started older lived a long time, with or without restriction, and their controls ate and weighed less than the Wisconsin controls; and the control males that started young weighed less than those in Wisconsin, although they ate the same. In those that started very young, no improvement in survival was seen, and the authors suggest it may have been a risk for some, although their data weren't yet final.10 In other words, the result depended heavily on what it was compared with.

How to read a headline about longevity

Part of the problem starts before the news story. In press releases from UK universities about studies in animals, cells or simulations, when the release extrapolated the results to people beyond what the scientific paper said, 86% of the news stories did so too, compared with 10% when the release didn't; this is an association, not proof that the release causes the exaggeration.11 And in a sample of 623 papers on Alzheimer's disease done in mice, nearly seven in ten news stories (1,067 of 1,535) didn't mention mice in the headline, especially when the scientific paper itself didn't name them in its title.12

Five questions to ask of a news story. They aren't a published rule but a deduction from what these sources show:

  1. Who was it done in? Cells, worms, mice, monkeys or people. If the headline doesn't say, look for it in the text or in the study itself.
  2. What was measured? How long they lived, a disease, or a marker that changes sooner. These aren't the same.
  3. What was it compared with? As with the monkeys, the control group can change the result.
  4. How much did it change, in absolute numbers? A “50% lower risk” can be a lot or very little depending on the starting risk; we explain this in how to read relative and absolute risk.
  5. Is it a single study, or has it been repeated? Even in mice, what works in one laboratory doesn't always work in another.5

If a news story makes you think about changing a habit, the useful thing is to turn it into a specific question about you and test it calmly, as we explain in how to turn a wellness data point into a question and a test. If it's about taking a drug or a supplement, that's a conversation to have with a health professional.

Studies in cells and animals are the beginning of many questions, not the answer to yours. A result in mice tells you what happens in mice; to know whether something extends human life you need studies in people, and measuring how long people live means following them for decades. In the meantime, it helps to read every headline with the same questions: in whom, what was measured, compared with what, and by how much.

Keep reading

Want to turn information into action?

naro helps you prioritize 2–3 high-potential habits and sustain them over time, with context and follow-up.

Explore the appWhat is naro

Sources

Each small number in the text leads to the source that statement rests on. Practical advice without a number is a deduction from these sources, and examples with a person's name are made up. It is general information: it does not replace an assessment by a health professional. How we make this content

Sources consulted up to 28 October 2026.

  1. Review of systematic reviews. Ineichen BV, Furrer E, Grüninger SL, Zürrer WE, Macleod MR (2024). Analysis of animal-to-human translation shows that only 5% of animal-tested therapeutic interventions obtain regulatory approval for human applications. PLOS Biology. 122 reviews, 367 therapies for 54 diseases. These are disease treatments, not interventions to extend life.
  2. Systematic review (research letter). Hackam DG, Redelmeier DA (2006). Translation of research evidence from animals to humans. JAMA. 76 highly cited animal studies, from 1980 to 2000; being the most visible, they may have reached people more easily than the rest.
  3. Methodological article. Guyatt GH, Oxman AD, Vist GE, et al. (2008). GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. How many guidelines rate confidence in the evidence; it doesn't discuss animal studies.
  4. Experimental study in mice. Harrison DE, Strong R, Sharp ZD, et al. (2009). Rapamycin fed late in life extends lifespan in genetically heterogeneous mice. Nature. Genetically diverse mice, in three laboratories. It doesn't study people.
  5. Experimental study in mice. Harrison DE, Strong R, Allison DB, et al. (2014). Acarbose, 17-α-estradiol, and nordihydroguaiaretic acid extend mouse lifespan preferentially in males. Aging Cell. Differences by sex and by laboratory within the same program. It doesn't study people.
  6. Experimental study in mice. Strong R, Miller RA, Astle CM, et al. (2013). Evaluation of Resveratrol, Green Tea Extract, Curcumin, Oxaloacetic Acid, and Medium-Chain Triglyceride Oil on Life Span of Genetically Heterogeneous Mice. The Journals of Gerontology: Series A. Negative results at the doses tested; it says nothing about other doses or about people.
  7. Perspective article. de Magalhães JP (2014). Why genes extending lifespan in model organisms have not been consistently associated with human longevity and what it means to translation research. Cell Cycle. Puts forward hypotheses based on other studies; it isn't a systematic review.
  8. Experimental study in monkeys. Colman RJ, Anderson RM, Johnson SC, et al. (2009). Caloric restriction delays disease onset and mortality in rhesus monkeys. Science. The University of Wisconsin study, after 20 years of follow-up. We could only read the abstract.
  9. Experimental study in monkeys. Mattison JA, Roth GS, Beasley TM, et al. (2012). Impact of caloric restriction on health and survival in rhesus monkeys from the NIA study. Nature. The study by the US National Institute on Aging. We could only read the abstract.
  10. Joint analysis of two studies in monkeys. Mattison JA, Colman RJ, Beasley TM, et al. (2017). Caloric restriction improves health and survival of rhesus monkeys. Nature Communications. 197 rhesus macaques; some were still alive when the data were analyzed, so part of the survival curves was incomplete.
  11. Observational study. Sumner P, Vivian-Griffiths S, Boivin J, et al. (2014). The association between exaggeration in health related science news and academic press releases: retrospective observational study. BMJ. Press releases from 20 UK universities in 2011 and the news stories they generated.
  12. Observational study. Triunfol M, Gouveia FC (2021). What's not in the news headlines or titles of Alzheimer disease articles? #InMice. PLOS Biology. Papers from 2018 and 2019 on Alzheimer's disease in mice; a single field of research.

← The whole library