^'Scientists discover Easter Island 'fountain of youth' drug that can extend life by 10 years," shouted a recent newspaper headline. "Coffee may 'reverse' Alzheimer's," said another. Amazing and shocking stuff, but there's a caveat – the research that fuelled the stories was done on mice.

Mus musculus is the most experimented-on creature in the history of humanity and you can bet that any modern pharmaceutical medical treatment or basic understanding of disease or genetics has involved working on a mouse at some point. Around 85% of all animals used in experiments are rodents and the vast majority of those are mice.

And for good reason. "Mice are used because they're the smallest and one of the easiest mammals to study in a laboratory setting – they breed quickly and are good enough for many types of study," says Simon Festing, chief executive of the pro-research charity Understanding Animal Research.

"While there are differences, we know that the main biological body systems work in the same way in all mammals. The reproductive, endocrine and cardiovascular and the central nervous systems all have a very similar structure and function," he says. "Mice share over 90% of their genes with humans."

Nevertheless, using a mouse can never tell scientists everything they need to know. A result on a mouse is an interesting lead but only replication in a higher animal, such as a dog or a monkey, pushes it closer to becoming a reality for people. In the case of the Easter Island elixir (a drug called rapamycin), reports suggested that the anti-ageing pill made from chemicals found on the islands had extended the life of mice by up to 38% – but the researchers warned that humans should not think about using the drug to extend life because it suppresses immunity.

And the Alzheimer's study, published in the Journal of Alzheimer's Disease, showed that caffeine could slow down the build-up of protein plaques, which are the signature of the disease and cause the damage to the brain. The mice were given the equivalent of five cups of coffee per day, containing around 500mg of caffeine, and showed almost a 50% reduction in the levels of the protein plaques in their brains after two months. But the scientists cautioned that, though caffeine was a relatively safe drug, there was no indication yet about the amounts of the chemical that would act successfully against Alzheimer's in humans. And pregnant women and people with high blood pressure should certainly avoid upping their caffeine intake.

There are several reasons why results on mice have problems translating directly to humans. When researching whether a drug works, doses on mice are sometimes much higher than anything that a doctor could safely use on a patient, even allowing for adjustments of metabolic rate and size. "Because your primary concern in the animal experiments is to demonstrate an effective treatment, you will dose higher than you would in a human," says Dominic Wells, head of the department of cellular and molecular neuroscience at Imperial College London.

"So if you hear a story that a mouse has been cured of this or that, you need to take that with a big pinch of salt because we would almost certainly not be allowed to take the same sort of dose rate straight into a human. We'd need to make a significant reduction to test for safety before we could consider upping the dose." Trials using animals tend to focus on a single question – efficacy or toxicity, for example. But to make something suitable for humans requires the management of side-effects, and this might take years to tackle.

In 2006, Daniel Hackam of the University of Toronto looked at how many animal-based experiments had been later verified by successful human trials. Out of 76 studies published between 1980 and 2000, 28 were successfully replicated in human randomised trials, 14 were contradicted in trials, and 34 remained untested. In a letter published in the Journal of the American Medical Association, Hackam wrote: "Patients and physicians should remain cautious about extrapolating the findings of prominent animal research to the care of human disease . . . poor replication of even high-quality animal studies should be expected by those who conduct clinical research."

There are also physiological limitations to experimenting on mice. "We can genetically modify, by a single intramuscular injection, a whole muscle in a mouse," says Wells. "If we try to do that in a person it just doesn't work because the spread of the agent we inject is maybe 4-5mm – the size of a mouse muscle." And in other areas, mice are not sophisticated enough to model humans. Neurologically, says Wells, "mice are wired in a different way. For us, the visual sense is incredibly important. If I showed you a blind mouse and a mouse with perfect eyesight in a cage, you wouldn't be able to tell the difference because mice rely a lot on smell and touch."

None of this should put a negative spin, however, on the importance of mice in research. So far, 26 Nobel prizes have gone to discoveries where research on mice has been key, including work on vitamins, the discovery of penicillin, the development of numerous vaccines and understanding the role of viruses.

At the DNA level, mice are increasingly useful in understanding how individual genes work and why they might go wrong. "We share the gene for the mouse tail and, while clearly we don't express that, it can tell us something about how gene expression differs between mice and humans," says Festing. "Adjacent to the tail gene is one for a certain type of childhood deafness that you get in both mice and humans. We've now been able to identify the cause of that deafness, which is actually of great benefit to the patients even though we haven't yet translated that into a cure."

Some recent advances

Breast cancer

A commonly used blood pressure drug seems to reverse the effect of a gene that has been implicated in up to a fifth of all breast cancers. Recent experiments on mice found that tumours where the AGTR1 gene was overactive shrank by nearly a third when a drug called Losartan was administered.

Stem cells

Experiments on mice have been critical in finding new sources (ie alternatives to embryos) of stem cells, both for experiments and medical applications. When scientists discovered a new type of stem cell that can be made by re-programming skin cells, they needed to know whether it behaved like genuine stem cells. Experiments on mice have been used to demonstrate that they do - most recently these new stem cells were used to successfully clone mice.

Blindness

A drug made from mouse anti- bodies has been found to slow the progression of a form of age-related macular degeneration, a major cause of blindness. In the disease, excessive amounts of blood vessels grow at the back of the eye, which then leak and cause damage to healthy tissue. The drug Lucentis inhibits a chemical messenger required for the eye to grow new blood vessels.

Gene therapy

Mice have been used to develop the next generation of medical treatments. Using hollow nanometre-sized particles of silver to deliver DNA into cells of mice, scientists have managed to reverse some of the symptoms of haemophilia. Gene therapy such as this can compensate for the effects of diseases caused by mutated DNA by delivering a "correct" copy of the genetic material to a site in the body.

Autism

In June, scientists found further evidence that more than one gene is involved in people with genetic-based autism. They found two genes in mice that, when mutated, caused autism-like symptoms. The genes interacted to affect brain growth and sociability.

Sources: understandinganimalresearch.org.uk; pro-test.org.uk