"An 'exercise pill' mimicking the effects of a gym workout could prevent Alzheimer's disease," the Daily Mirror reports.
The "pill" is actually a reference to a protein called irisin. Irisin has been dubbed the "exercise hormone" because previous research found it's released from muscles in response to physical activity.
Researchers wanted to see if irisin, or the lack of it, had any role or impact in Alzheimer's disease.
This is because irisin has previously been found in the area of the brain called the hippocampus, which is involved with learning and memory. Both these cognitive functions are adversely affected by Alzheimer's.
The researchers looked at post-mortem brain specimens from elderly adults.
They found people who'd had advanced Alzheimer's had lower levels of irisin in the hippocampus than people who'd had earlier stage disease or normal brain function.
They then carried out experiments in mice bred to have an Alzheimer's-like condition.
They found blocking irisin production worsened the mice's memory and learning.
Boosting irisin levels through a daily swimming programme helped to improve memory and learning.
These are interesting findings that may help improve our understanding of Alzheimer's.
But at this stage there's no evidence that exercise will directly prevent Alzheimer's, or restore brain function in those with the disease. Further studies in people with Alzheimer's are needed.
That said, the study supports the idea that physical activity may reduce the risk of many long-term diseases and may preserve brain health as we age.
Find out more about the many benefits of exercise
The study was conducted by researchers from the Federal University of Rio de Janeiro in Brazil and other institutions in the US and Canada.
The work received funding from numerous organisations, including the Alzheimer Society of Canada and the Weston Brain Institute.
It was published in the peer-reviewed journal, Nature Medicine.
The study was reported accurately in the UK media. But some of the headlines could have made it clearer that this was only a very early-stage laboratory study that mainly involved mice.
This laboratory study involved an animal model of Alzheimer's disease and analysis of brain samples from humans.
The brain is known to be a target for various hormones that may help brain cell survival and stimulate the formation of new connections between nerve cells.
Failure of these types of hormonal activities has been linked with brain disorders, particularly Alzheimer's.
This research centres on the hormone irisin, a protein that's released from muscle cells as a result of physical activity.
Irisin has been detected in the area of the brain called the hippocampus, which is known to be involved with learning and memory.
It's thought irisin released in response to exercise could have a protective role in Alzheimer's.
This early-stage laboratory study may give us an indication of the biological process involved in preserving learning and memory, but it can't prove that exercise protects against Alzheimer's in humans.
This laboratory study involved mice bred to have a condition like Alzheimer's.
It also involved post-mortem brain samples from around 20 elderly adults, some whom had had early- or late-stage Alzheimer's, plus age-matched controls from people who had no history of Alzheimer's.
The researchers then carried out tests to see if irisin was present in the hippocampus of the mouse and human brains. They compared the levels in people with and without Alzheimer's.
They then looked at whether the build-up of amyloid plaques (abnormal clumps of protein characteristic of Alzheimer's) was associated with irisin levels.
They also gave mice a virus that "knocks out" irisin to see what effect this had on their memory and behaviour when exposed to various tests, such as a water maze test.
They compared this effect in both normal mice and those bred to have the Alzheimer's-like condition.
The researchers then looked at the effects of restoring irisin in mice.
The study confirmed that irisin was present in the hippocampus of the mouse and human brains.
They found levels were lower in mice with Alzheimer's-type disease.
They were also lower in humans with late-stage Alzheimer's compared with controls or those with earlier stage disease.
They found higher levels of amyloid in humans and rodent brains were associated with reduced irisin levels.
Knocking out irisin did not affect memory or behaviour in normal mice. But it did affect test performance in the Alzheimer's mice.
They found the absence of irisin affected the ability of the mouse brains to form new nerve connections.
The researchers confirmed that boosting irisin levels in the Alzheimer's mice restored their memory defects and the ability to form new nerve connections.
They also showed that giving Alzheimer's mice an "exercise regime" of daily swimming boosted their hippocampal irisin levels.
The researchers said their findings "place FNDC5/irisin as a novel agent capable of opposing [nerve connection] failure and memory impairment in Alzheimer's disease".
They suggested that exercise could be a way to increase hippocampal irisin levels in people at risk of Alzheimer's or in those who already have cognitive impairment.
This is an interesting study that seems to show that the memory and learning region of the brain in those with Alzheimer's has lower levels of the protein irisin.
But what this actually means is an entirely different question. It could be that low physical activity levels during the person's lifetime has led to lower levels of irisin in the brain, and that this has led to the development of Alzheimer's.
An alternative explanation could be that brain changes that occur during Alzheimer's prevent irisin accumulating to the same extent.
There's also the possibility that people with advanced Alzheimer's do less activity because of their illness and this is why they have low levels of irisin.
This study alone doesn't explain the potential role of irisin.
The research did find that restoring irisin in Alzheimer's mice, including by daily swimming, can boost memory and nerve connections.
But we don't know that the same effects would be seen if people with Alzheimer's were given an exercise programme, for example.
We also don't know if any attempt to increase irisin levels through drug-based treatments (as some in the UK media suggested) would be either effective or safe, as no studies have been carried out in people at this stage.
Nevertheless, this early-stage study does support our understanding that regular physical activity has many health benefits and may reduce the risk of many long-term diseases.
The causes of Alzheimer's remain poorly understood, but a healthy lifestyle that includes a balanced diet, regular exercise, not smoking and only drinking alcohol in moderation may all help preserve brain health as we age.