The thing they want it to do sounds almost embarrassingly basic for 2026: draw a complete chemical map of the lunar surface. We don’t have one. After Apollo, after a parade of orbiters from half a dozen nations, the Moon’s elemental geography is still a patchwork of well-surveyed scraps and enormous blanks.

The reason comes down to how this kind of mapping works. Fire enough solar X-rays at lunar rock and the atoms in it fluoresce, each element coughing back X-rays at its own signature energy. Catch those and you can read off what the ground is made of, oxygen and iron and silicon and the rest, no shovel required. The catch is that you are entirely at the mercy of the Sun. No flare, no signal. And the Moon’s poles, the very places everyone now wants to land, sit at such a glancing angle to the Sun that the X-rays arrive feeble and slanting, barely enough to register.

So the polar regions, scientifically the most interesting real estate in the inner solar system right now, are precisely where the old technique goes blind.

Why Nobody Had Done This Already

Earlier missions did try. Apollo 15 and 16 mapped roughly a tenth of the surface. India’s Chandrayaan-2 managed a respectable 12.5-kilometre resolution but left gaps wherever the Sun stayed quiet, and China’s Chang’E-2 swept across about 65 per cent of the Moon yet couldn’t cleanly separate the signals of magnesium, aluminium and silicon, which sit awkwardly close together in energy. The instruments kept running into the same two walls: solar flares that didn’t oblige, and detectors that slowly cooked in space radiation until their readings smeared.

There was a third problem, more mundane and in some ways more stubborn. To stare at a wide swathe of ground during one of those rare, generous flares, you really want a telescope. But the X-ray spectrometers flown so far couldn’t carry one, because the only way they had to focus was a mechanical collimator, essentially a heavy honeycomb baffle, and bolting a proper telescope on top would have made the payload hopelessly bulky.

This is where the lobster eye earns its keep. The optics that Airi Toida, Yuichiro Ezoe and their colleagues at Tokyo Metropolitan University have adapted were not designed for the Moon at all. They were built for a small satellite mission called GEO-X, meant to photograph the faint X-ray glow of Earth’s own magnetosphere. The whole imaging unit, optics and sensor and filter together, is about the size of three stacked drink cartons and weighs under ten kilograms. You could, in principle, tuck it onto a spacecraft already going somewhere else.

What Two Years of Patience Buys You

The new work doesn’t fly the thing. Instead the team built a detailed numerical model, feeding in the telescope’s real measured quirks, its efficiency, its field of view, the way its sensitivity tails off toward the edges, and then simulated what a satellite in a polar orbit around the Moon would actually see over a realistic mission. They assumed a fairly ordinary Sun, around 300 flares of various sizes a year, which is about what our star tends to manage. Then they asked the obvious question: how long until you’ve gathered enough light to trust the map? The radiation worry, at least, looks manageable. In testing under conditions harsher than lunar orbit, the sensor’s energy resolution drifted by less than 50 electronvolts even after a punishing dose, which for this sort of work is reassuringly little.

The answer, for a single telescope: about two years to map five elements, oxygen, iron, magnesium, aluminium and silicon, across the entire surface at a resolution of 70 kilometres to a side. Iron and oxygen come fast, inside a few hundred days; the others take patience. Not a fine-grained map, but a complete one, which is the part nobody has managed.

And because each unit is so light, you needn’t stop at one. Pack 25 of them into a five-by-five array and the field of view balloons, letting the spacecraft drop to a lower orbit and tighten the grid to 30 kilometres while finishing in roughly a year. That version even teases out sodium, a notoriously shy element, within two years.

There are honest limits, and the paper doesn’t hide them. The heavier elements, calcium and titanium and the like, would need more than a century to map globally this way; you’d only ever catch them locally, during the strongest flares. And those 25 detectors are hungry. The cameras alone would draw at least 25 watts, which on a small spacecraft is not nothing, and that is before you’ve powered anything else. Whether the sums close is a question for mission planners, not models.

Still, the appeal is hard to miss with NASA’s Artemis programme circling back to the Moon and its Gateway station meant to hang in lunar orbit. A complete chemical atlas would do more than satisfy curiosity about how the Moon cooled and differentiated billions of years ago. It would tell future crews what the ground beneath a candidate landing site is actually made of before anyone commits to going. The poles especially, where the water and the ambitions both seem to be.

For now it lives in a computer, an instrument that hasn’t flown answering a question no one has yet funded it to ask. But the hard part, the optics, already exists and has the radiation scars to prove it. Sometimes the breakthrough isn’t a new idea so much as noticing that the right tool was sitting in the next room, built for something else entirely.

Source: Toida et al., Earth, Planets and Space (2026)

Frequently Asked Questions

Why can’t we just photograph the Moon to figure out what it’s made of?

Ordinary cameras see reflected sunlight, which tells you about colour and texture but not reliably about chemistry, especially for lighter elements like magnesium and aluminium. X-ray fluorescence works differently: solar X-rays make the atoms in the surface emit their own X-rays at energies unique to each element, so you’re reading a direct chemical fingerprint rather than guessing from appearance. The trouble is you need the Sun to cooperate with a flare first.

What does a lobster’s eye have to do with any of this?

X-rays won’t focus through a normal curved lens; they pass straight through or get absorbed. Lobster-eye optics get around this with a grid of tiny square channels that reflect X-rays off their inner walls at shallow angles, bending them toward a focus the way the real animal’s eye gathers dim light underwater. Mimicking that structure in etched silicon is what makes the telescope light enough to fly on a small satellite.

If the map is so useful, why hasn’t anyone made one yet?

Every previous attempt hit the same obstacles: solar flares are unpredictable, so coverage came in patches, and space radiation gradually degraded the detectors. On top of that, older X-ray instruments were too heavy to carry a real telescope, which limited how much ground they could survey during the brief windows when the Sun did flare. Getting past all three at once is what’s new here.

Could this actually help astronauts on future Moon missions?

That’s a large part of the motivation, given Artemis and the planned Gateway station in lunar orbit. A full chemical map would let mission planners know what the surface is made of at candidate landing sites before committing, particularly around the poles where future crews are headed. It’s the difference between scouting the whole neighbourhood and knowing only a few scattered streets.

The compounds in question are flavanols, a family of plant chemicals found in tea, cocoa, berries, apples and certain beans. They are not vitamins, and your body does not strictly need them to survive, which is rather the point of all the fuss.

Here is where it gets interesting. In the COSMOS trial, the biggest randomised study of these polyphenols ever run, a daily dose of 500 milligrams of flavanols cut deaths from cardiovascular disease by 27 per cent. That is a striking number for something you can, in principle, eat. The obvious assumption, the one that has underpinned decades of public-health advice, is that anyone dutifully eating their fruit and veg must already be getting plenty. The new work, published in the journal Food & Function by scientists from Reading, Harvard, UC Davis and the confectionery-and-nutrition firm Mars, set out to check whether that assumption holds.

It doesn’t. Across two large cohorts, one American and one British, fewer than one in five people reached the 500-milligram mark.

And these were not people slacking on their greens. The researchers leaned on two urinary biomarkers rather than the notoriously slippery food diaries, measuring molecules with tongue-twisting names that the body produces only after flavanols pass through it. Even among participants who hit the recommended five portions a day and scored well on overall diet quality, only around a fifth crossed the threshold linked to a healthier heart.

“Most people assume that eating plenty of fruit and vegetables covers this, but what this research shows is that the specific choices you make matter far more than the total amount,” says Javier Ottaviani, the paper’s lead author. A handful of blackberries, a whole apple, a cup of green tea alongside lunch, he suggests, could shift the dial in a way that a banana and a carrot simply won’t.

Not All Five-a-Days Are Created Equal

The arithmetic, once you see it, is faintly absurd. A punnet of plums delivers something like 450 milligrams of flavanols in one go; a cup of green tea, around 200. A medium apple with the skin on manages perhaps 110, blueberries a modest 80. The trouble is that flavanol content swings wildly even within a single fruit. The team notes that levels of one key flavanol, (-)-epicatechin, can vary more than tenfold between apples of the very same variety, depending on the cultivar, the weather, when it was picked and how it was stored. Put bluntly, it could take anywhere from two apples to twenty-nine to land the dose tested in COSMOS. So your apple a day might be doing the job, or it might be doing almost nothing, and you would have no way of telling them apart.

What gives the figures their bite is the way they were collected. Rather than trusting people to recall what they ate, which study after study has shown to be optimistic at best, the team measured flavanol breakdown products directly in urine, using two markers with slightly different half-lives so they captured both the last couple of hours and the wider day. The thresholds were even set generously, deliberately tilted to overcount the people clearing the bar. Which means the true picture is, if anything, a little bleaker than the one in the paper.

The British data threw up a wrinkle that is harder to wave away. In the UK cohort, the people who adhered most closely to national dietary guidelines were actually the least likely to hit the flavanol target.

Why? Probably tea, oddly enough. Tea is one of Britain’s great flavanol sources, and the keenest guideline-followers were not necessarily the keenest tea-drinkers. Even so, the gap between best-practice eating and the heart-protective dose was real, and it points to the same conclusion on both sides of the Atlantic.

Whether We Need New Numbers

So should the advice change? The researchers are careful not to tear up five-a-day, a message that earns its keep on plenty of other grounds. But they argue the headline number is no longer enough on its own. An expert panel in the US has already floated a target of 400 to 600 milligrams of flavanols a day for cardiometabolic health, and this study hands that idea a good deal of ammunition.

“Five-a-day is the right message, but we may need to think more carefully about which five,” says Gunter Kuhnle of the University of Reading. Get that right, and the humble blackberry starts to look less like a snack and more like a small, deliberate act of cardiology.

Read the full study in Food & Function (DOI: 10.1039/D6FO00867D)

Frequently Asked Questions

Why does it matter which fruit and vegetables I eat, if I’m already hitting five-a-day?

Because the heart-protective compounds called flavanols are spread very unevenly across produce. A punnet of plums or a cup of green tea can deliver hundreds of milligrams, while a banana or a carrot contributes almost none. Hitting your five portions tells you nothing about whether you reached the flavanol dose linked to lower cardiovascular risk, which is where the food you pick becomes the deciding factor.

Is it true that healthy eaters can still fall short on flavanols?

Surprisingly, yes. In the large new analysis, even people who met dietary guidelines and scored well on overall diet quality mostly failed to reach the 500-milligram threshold. In the UK group, the closest adherents to official advice were actually the least likely to hit it, partly a quirk of who drinks the most tea.

How much of a difference do flavanols actually make to heart health?

In the COSMOS trial, the largest study of its kind, a daily intake of 500 milligrams of flavanols was associated with a 27 per cent reduction in deaths from cardiovascular disease. That is a meaningful effect for a compound found in everyday foods, which is exactly why researchers are pushing to work out how people can reliably reach that level.

What’s the easiest way to get more flavanols into my day?

Reach for the high-flavanol options: berries such as blackberries and plums, apples eaten with the skin on, broad beans, and tea, particularly green tea. A single cup of tea or a punnet of berries can outweigh several low-flavanol portions combined, though natural variation means no single food is a guaranteed fix.

The enzyme is GRK2, and on a normal day it does unglamorous work. It helps cells read incoming signals correctly, in the heart, in the brain, pretty much everywhere.

But Quitterer, Professor of Molecular Pharmacology at ETH Zurich, and her team found that in Alzheimer’s brains a corrupted version of GRK2 shows up in alarming quantities. Cells can switch the enzyme off through their own metabolism, and that switched-off form does something nasty: it clumps. In aged mice engineered to develop the disease, roughly 64 per cent of the GRK2 in one memory-critical brain region had collapsed into these aggregates, against less than 9 per cent in healthy animals. The clumps drift toward the mitochondria, the tiny power plants inside every neuron, and settle there like grit in an engine.

What happens next is the part that should worry anyone who has watched a relative fade.

“The GRK2 aggregates block the pores of the mitochondria, reducing the amount of energy they can supply and leading to a situation of stress inside the cells,” says Quitterer. Starved and stressed, the neurons start producing more amyloid beta, the protein fragment that has dominated Alzheimer’s research for decades. And here is the cruel twist. That extra amyloid stresses the cells further, which produces still more broken GRK2, which makes more amyloid. Round and round it goes.

The team traced exactly what the rogue enzyme grabs hold of: a small mitochondrial protein called TOMM6, whose ordinary job is helping to keep the power plant’s import machinery assembled. Trapped by the aggregates, TOMM6 stops doing that job, and the mitochondria suffer for it.

Breaking the Circle

So the question Quitterer’s group set themselves was blunt: could you persuade GRK2 to stay in its functional form? They built a series of small molecules and tested them in cell cultures and in mice, and one candidate, the now-familiar Compound 10, did the trick. It nudges the equilibrium back toward healthy, monomeric GRK2 and away from the clumping form. With the enzyme behaving again the mitochondria recovered, amyloid deposits shrank, and the neurons that would otherwise have died simply carried on living. Mice given the compound orally from middle age survived longer than untreated ones. The molecule slips across the blood-brain barrier readily, which is no small thing for a drug aimed at the brain, and across more than 40 other pharmacological targets it left no fingerprints, a reassuring sign on the safety front.

There were odder effects, too. Treated animals showed better heart function and, charmingly, fewer grey hairs in old age.

None of this came quickly, and Quitterer is candid about why. “It took so long simply because everything takes so long in Alzheimer’s research,” she says. Because the disease is one of aging, the experiments demand old mice, animals of perhaps one and a half to two years, and each round of work eats another eighteen months or more before it yields anything you can build on. “It’s all a great deal slower than in cancer research, for example.”

A Different Door

What makes the finding interesting is not that it promises a cure, because it does not. “Alzheimer’s is a very complex disease,” Quitterer notes, and today’s drugs at best delay the slide by a matter of months. The value lies in the direction of attack. “That’s why it’s so important that we’ve now identified a new target protein in the form of GRK2, as well as an active ingredient that operates via GRK2 and therefore via a different mechanism than existing Alzheimer’s drugs,” she says. A drug that works by a wholly different route might, one day, be paired with the ones we already have.

The caveats are real and the team does not hide them. The roots of the project reach back to brain tissue collected during tumour surgery at the Ain Shams University Hospital in Cairo, and the human side of the evidence rests on a mere handful of patients, a limitation Quitterer’s group flags plainly. Everything else, for now, lives in mice and cell dishes, and the long graveyard of Alzheimer’s drugs that shone in rodents and failed in people is a warning nobody in the field forgets.

Still, there is something compelling about a target that sits at the crossing point of amyloid, faulty mitochondria and the broader machinery of aging, rather than chasing any one of them alone. ETH Zurich has filed for a patent and is now hunting for a company willing to carry Compound 10 toward an actual drug. Whether it gets there is anyone’s guess. But after two decades of patient acquaintance, Quitterer has at least handed the field a new door to try.

Full study: Cell Reports Medicine, DOI 10.1016/j.xcrm.2026.102707

Frequently Asked Questions

Why would an enzyme the body needs end up causing harm?

GRK2 is essential in its normal, working form, but cells can chemically switch it off, and in Alzheimer’s brains that inactive version builds up and clumps together. The clumps settle on mitochondria and choke their energy supply, which sets off a damaging cascade. So the trouble is not the enzyme itself but a corrupted form of it accumulating where it should not.

How is this different from the amyloid-targeting drugs we already have?

Most current Alzheimer’s treatments aim directly at amyloid beta or its plaques. This approach goes upstream, stabilising GRK2 so the chain reaction that pumps out amyloid never gets going in the first place. Because it works through an entirely separate mechanism, researchers think it could one day be combined with existing drugs rather than replacing them.

Is Compound 10 something patients could take soon?

Not yet, and possibly not for years. The results so far come from mice and laboratory cell cultures, with only a small number of human tissue samples backing them up. The basic research is finished and a patent is filed, but the molecule still needs a commercial partner and the full gauntlet of human trials before anyone could call it a medicine.

Why did the research take almost twenty years?

Alzheimer’s is a disease of aging, so the experiments need elderly mice, and growing them plus running each study can take well over a year apiece. As the lead researcher puts it, the whole field moves far more slowly than something like cancer research. That glacial pace is part of why genuinely new drug targets are so rare.

That patch is the handiwork of Tae-Woo Lee and his team at Seoul National University, with collaborators at Stanford, and it tackles a problem that has dogged a curious class of devices for two decades. The thing they have built is an organic light-emitting transistor, and until now those have been maddeningly difficult to coax into doing anything useful.

The appeal is easy to see. A transistor switches and processes signals; a light-emitting diode glows. Fold both jobs into a single chunk of light-emitting plastic and you get a component that can sense, compute and display all at once, no separate parts to wire together. Wearable electronics, the kind people imagine eventually living on or even inside the skin, want exactly that: sensing, processing, memory and a readout, all in one soft, flexible package. The catch has always been the voltage.

Conventional versions of these transistors are thirsty. Because the electrodes sit far apart and electrons struggle to get into the plastic, the older field-effect designs demand somewhere between 80 and 180 volts to light up, which is plainly hopeless for anything worn against a body. Switch to an electrochemical design, where ions in a gel do some of the heavy lifting, and the voltage drops, but you still need more than 3.5 volts, and the patch of light that results is narrow and tends to wander about as the device runs.

Coaxing Electrons In Through the Back Door

Lee’s group went after the electron problem sideways. The trouble in a polymer like MEH-PPV, the orange-red emitter they used, is that holes flow easily but electrons barely get in at all, so the two never meet to produce light. Their fix was to blend a humble ingredient into the plastic, an ion transport enhancer (in this case a common surfactant, the sort of molecule you might find in a detergent), which lets positively charged ions shuffle through the film far more freely. Those ions pile up at one electrode and form what is called an electric double layer, a wafer-thin sheet of charge that effectively props the door open for electrons. No aggressive chemical doping required.

The upshot is light at startlingly low voltage. The device glows even when the drain voltage sits below the polymer’s own bandgap potential, around 2.17 volts, a threshold that on paper ought to be the floor for emission. In practice the whole thing runs happily on two 1.5-volt batteries, the kind rattling around in a kitchen drawer.

There is a subtler payoff buried in the mechanism, and it may matter more than the voltage. In older electrochemical designs the glowing region forms where a moving junction of positive and negative doping happens to meet, and that meeting point drifts, which makes the emission spot small and twitchy and hard to control. Because Lee’s approach skips the doping front entirely and pins the charge layer to the drain electrode, the light-emitting zone stays put. The researchers measured a recombination zone up to 267 micrometres wide, several times broader than the sub-75-micrometre flickers managed before, and crucially it does not slide around when the gate voltage changes. A wide, stable stripe of light is the difference between a lab curiosity and something you could actually build a display from. They went on to make working arrays, a 4-by-4 and then a 10-by-10 grid of a hundred glowing pixels, and bent and twisted the flexible versions without killing them.

None of this means a rollable plastic television is arriving next year. The brightness, while respectable for the genre at up to 826 candela per square metre, is modest next to a commercial screen, the pixel density is low, and the lifetime, though much improved at thousands of seconds of stable operation, is still measured in hours rather than years.

And the resolution, a couple of pixels per inch in the demonstration arrays, is nowhere near what a phone display needs. This is early-stage stuff, proof that a principle works rather than a product you could buy.

A Patch That Feels Pain

Where it gets genuinely interesting is what the team did with the memory. Because ions linger in the film after a stimulus, the transistor remembers, brightening more under repeated or sustained prods and holding that brightness for a while, much as a nerve grows more responsive to insistent pressure. The researchers wired one into a small, battery-powered, flexible system they call a stand-alone neuromorphic display, designed to mimic the way biological pain works. Gentle contact gets filtered out below a threshold; only a genuinely harmful poke crosses it and triggers a visible flash of light. The pitch is for people with pain-insensitivity disorders, who cannot feel when they are being hurt, and the obvious imagined home for such a thing is artificial skin. “This work is particularly meaningful in that it demonstrates that all functions can be integrated within a single semiconductor device, without the need to separately fabricate and connect processing, memory, and display units,” says Lee.

Whether any of this reaches a clinic or a wrist is, of course, an open question, and the leap from a centimetre-scale patch in a Seoul lab to a certified medical device is a long one. But the underlying idea, that you might not need separate chips for thinking, remembering and showing, has a tidy logic to it that biology arrived at long ago.

For now the group is looking skinward. “Going forward, we plan to further develop this technology into an on-skin semiconductor platform applicable to intelligent artificial skin and wearable healthcare,” says Lee. A glowing scrap of plastic that knows when it has been hurt is a strange object to hold in your head, and stranger still to imagine wearing.

Source: Kim et al., Nature Materials (2026), DOI 10.1038/s41563-026-02613-7

Frequently Asked Questions

How can a transistor and a light source be the same thing?

A transistor controls the flow of electric charge, while a light-emitting diode glows when positive and negative charges meet inside it. If the transistor’s channel is made from a light-emitting plastic, both jobs can happen in the same material: charges flow through it to switch and process signals, and where they recombine, the material glows. The hard part has always been getting electrons into the plastic without enormous voltages, which is what this work solves.

Why does running on two AA batteries matter so much?

Older light-emitting transistors needed anywhere from a few volts to well over a hundred, which rules them out for anything worn on the body. Getting stable, reasonably bright emission below 3.5 volts means the device can run on the same power you would use for a TV remote. That is the threshold at which “wearable” stops being a slogan and starts being plausible.

Is this ready to replace the screen on a phone?

Not remotely. The demonstration arrays hold a hundred pixels at a resolution of roughly two or three pixels per inch, the brightness is modest, and the operating lifetime is measured in hours. It is a proof of principle that the underlying mechanism works, not a finished display you could buy.

What does it mean for the patch to “feel pain”?

The device holds a faint memory of past stimulation because ions linger in its film, so it brightens more under repeated or sustained pressure and ignores gentle contact below a set threshold, much like a biological nerve. The team built a flexible, battery-powered version that flashes only when a poke is strong enough to count as harmful. The imagined use is artificial skin for people who cannot feel pain and so cannot tell when they are being injured.

That odd split state is the work of researchers at the University of Wisconsin-Madison, who have managed to coax one region of the brain into a sleep-like pattern while the rest stays online. The trick, reported this week in Nature Neuroscience, seems to hand that small patch some of the restoration it would normally only get from a proper night’s rest.

To understand why that matters, you have to know what the sleeping brain is actually up to. During non-REM sleep, which accounts for roughly 80 per cent of an adult’s nightly total, cortical neurons stop their waking chatter and start firing in synchrony: everyone on, everyone off, over and over, hundreds of times a minute. These are the slow waves you would see on an EEG. They are widely thought to be the moment when the brain takes stock of its connections, strengthening the junctions worth keeping, pruning the ones it does not need, and clearing space to learn again the next day.

So the question more or less asks itself. If those on/off rhythms are doing the restorative heavy lifting, could you simply install them by hand?

Chiara Cirelli and her colleagues had a head start. They had shown previously that sleep-deprived rats and humans slip into brief, sporadic patches of slow-wave activity even while awake, a phenomenon they call local sleep. The snag is that these episodes are too short and too scattered to do much good (and they can muddle your performance if they strike the wrong region at the wrong moment). What nobody had tried was making the pattern deliberate, sustained, and aimed at a chosen spot.

A patch of cortex, switched off to order

“What we’re essentially doing is forcing sleep in a local region of the brain,” says Cirelli. The team reached for optogenetics, the technique that uses light to switch genetically tweaked neurons on and off.

They tried two routes into the same destination. In one set of mice they used light to fire up a class of inhibitory cells called somatostatin interneurons, which act as a sort of master brake on the local circuit; in another they silenced the excitatory pyramidal neurons directly. Either way, for 30 minutes at the tail end of a five-hour sleep deprivation, one side of the cortex was driven through the slow rise and fall of induced off periods while the mouse stayed awake and busy. Then they let the animals sleep, and watched. On the stimulated side, slow-wave activity in that subsequent sleep was lower than on the untouched side. In plain terms, that bit of brain behaved as though it had less catching up to do. It needed less sleep, because in a sense it had already had some.

Here is the part that surprised me. You might assume the benefit comes simply from giving tired neurons a rest, from dialling the firing down. Some researchers had argued exactly that. But when the team used a different tool to clamp the overall firing rate down to the same low level, without the rhythmic alternation, the effect vanished. No drop in subsequent slow waves, no sign of relief. It was the on-and-off pattern itself that mattered, the switching, not the silence.

The molecules told the same story. After the awake stimulation, the treated cortex showed lower levels of certain AMPA receptors, the proteins that register the strength of a synapse, much as you would expect to find after a stretch of real sleep.

Memory rescued from a sleepless night

Then came the test that counts. Mice learned to tell two floor textures apart, a task that leans on sleep to bed the memory down. Some were allowed to sleep; some were kept awake for an hour; and some were kept awake but given the on/off stimulation across both sensory and motor cortex. The sleep-deprived animals that got nothing did poorly the next day. The ones that got the stimulation, despite missing their sleep, remembered roughly as well as the mice that had slept. The memory had been salvaged, in effect, without the sleep that normally carries it.

None of this means a gadget for skipping sleep is anywhere close. The work is in mice, the method involves light-delivering implants and genetic modification, and the brain-wide reset of a full night is almost certainly doing things that no local patch can replicate. Cirelli is more interested in whether the same effect might be reached gently, from outside the skull, with transcranial stimulation, and that is where she wants to take it next.

“This research further decodes why we sleep and how we learn, which brings us a step closer to understanding how to better prevent and treat cognitive decline,” says Amy Bany Adams of the US National Institute of Neurological Disorders and Stroke, which funded the work. The dolphins, who sleep one hemisphere at a time while the other keeps watch, worked this out long before we did. We are only now learning to ask the brain for the same favour, one small region at a time.

DOI / Source: https://doi.org/10.1038/s41593-026-02318-9

Frequently Asked Questions

Can this actually let people skip sleep?

Not any time soon, and possibly never in full. The study was done in mice using light-sensitive implants and genetic modification, and even then it only reset small patches of cortex rather than the whole brain. A full night’s sleep coordinates restoration across the entire brain in ways a local trick cannot match, so this is a window into how sleep works, not a replacement for it.

Why does the on-and-off pattern matter more than just resting the neurons?

That was the unexpected core of the finding. When researchers simply lowered the overall firing rate without the rhythmic switching, the restorative effect disappeared entirely. It seems the repeated transitions between firing and silence are what drive the brain to recalibrate its connections, which overturns the idea that mere neuronal rest is enough.

How does sleep recalibrate the brain’s connections in the first place?

During non-REM sleep, neurons fire in synchronised on/off cycles that show up as slow waves on an EEG. This is thought to be when the brain strengthens the synapses worth keeping, weakens the ones it does not need, and frees up capacity to learn again. The new work backs this up by showing that inducing those cycles lowers molecular markers of synaptic strength, just as real sleep does.

Could this ever help with conditions like cognitive decline?

That is the long-range hope rather than a current capability. The researchers and their funders frame the work as decoding why we sleep and how we learn, which could eventually inform treatments for memory and cognitive problems. The nearer-term goal is testing whether a non-invasive version, using stimulation through the skull, can produce a similar effect in humans.

For a small group of patients, that reading keeps coming back month after month, year after year, with no obvious reason. Clinicians call it nonsuppressible viremia, and it has long been a genuinely awkward problem.

It affects fewer than one in a hundred people on long-term therapy, but the consequences ripple outward: extra clinic visits, changes to medication, the quiet suspicion from some providers that the patient simply isn’t taking their pills. Now a team at Johns Hopkins, working with collaborators across the United States, Canada and Denmark, has pinned down what is actually going on in most of these cases. And the answer is reassuring in a way few expected.

The virus showing up in the blood is, for the most part, broken.

Writing in Nature Communications, the researchers looked at plasma from more than 50 people whose virus kept surfacing despite faultless treatment. In about 95 per cent of cases, the HIV genetic material floating in their blood carried a crippling flaw, a mutation or deletion in a stretch of RNA called the 5′ leader. This region acts as a kind of control panel for the virus, governing how it copies and packages itself. Damage it, and the virus can still be made, but it can’t infect anything.

A factory still running, making faulty goods

Here’s the wrinkle that makes this counterintuitive. Modern HIV drugs, which have been around since 1996, stop the virus from infecting fresh immune cells. What they can’t do is reach back into cells that were infected long ago and silence them. Most of those cells stay quiet. A few, expanded into clones over the years, keep churning out viral particles, and it turns out the genomes they’re churning out are duds.

“We know that these defective proviruses cannot infect new cells, but they are still clinically relevant,” says Francesco Simonetti, the senior author and an infectious diseases physician at Johns Hopkins. “Think of how many extra visits, extra drugs, extra costs and tests they’ve been causing.”

What surprised the team was just how lopsided the picture is. Inside the reservoir of infected CD4 cells, these 5′ leader defects are rare, barely over one per cent. Yet in the blood, the defective versions accounted for nearly all of the circulating virus, around 98 per cent in the cells they examined closely. The broken genomes were massively over-represented in the plasma, and the intact, dangerous ones were the ones being quietly cleared away. Simonetti’s reading of this is that over time on treatment, the proviruses that make functional virus get pruned by the immune system, while the defective ones slip past it and accumulate.

The damage, when they mapped it, clustered with surprising precision around a single spot, the major splice donor. It’s the site where the cell’s splicing machinery, guided by something called U1 snRNA, normally latches on. Knock out the right nucleotide there and the splicing falls apart, the virus loses its ability to replicate, but transcription carries on regardless. The cell keeps reading a broken blueprint and shipping out the useless product.

A test borrowed from cancer

Spotting all this the slow way, by sequencing virus one genome at a time, is expensive and fiddly. So the team built a shortcut. Their assay, named CLAWS, for Capturing 5′ Leader Anomalies Without Sequencing, runs on digital PCR and uses two molecular probes: one that always binds, and a second that only sticks if the splice site is intact. If the second probe goes dark, the virus is broken. The trick is borrowed straight from oncology, where similar “drop-off” tests hunt for cancer mutations in cell-free DNA.

It is cheap, fast, and sensitive enough to catch nine copies of virus in a millilitre of plasma. Used on samples taken soon after people started treatment, CLAWS picked up defective genomes emerging within about a month, then watched them take over. In people who’d been on therapy for two decades, the defective forms made up more than 95 per cent of what was left.

There is a catch, naturally. The assay was built almost entirely on subtype B virus, the dominant strain in wealthy countries, and it will need checking against the diverse subtypes that drive the epidemic globally before anyone rolls it out widely. It also can’t yet read residual virus down at the single-copy level, though the researchers reckon that’s within reach.

Still, the practical payoff could land soon. A clinician who can confirm that a stubborn viral load is just defective debris could skip the extra drugs and the anxious conversations. People could qualify for hip or knee replacements, organ transplants, or cure-focused clinical trials they might otherwise have been shut out of. “From a clinical perspective, this is important because people with HIV are taught that the absolute goal of their medication is to achieve undetectable viral load and they worry,” Simonetti says.

And the broken virus may yet have something to teach. The fact that defective genomes survive while functional ones are culled hints at a vulnerability, some difference in how the immune system sees the two. Pin that down, and it might point the way toward clearing the reservoir that has kept HIV out of reach of a cure for forty years.

Read the full study in Nature Communications

Frequently Asked Questions

If the virus showing up is broken, why does it still set off the test?

Standard viral load tests detect HIV genetic material, not whether that material can actually infect anyone. A defective provirus can still be transcribed and packaged into particles that show up as detectable RNA, even though the virus inside is incapable of replicating. That’s exactly why a detectable result has historically caused so much alarm: the test couldn’t tell functional virus from broken debris. A newer assay aims to close that gap.

Does a detectable viral load always mean someone’s treatment is failing?

Not necessarily. In the large majority of cases studied here, persistent low-level virus in adherent patients came from defective genomes, not from the virus actively replicating or developing drug resistance. That distinction matters enormously, because it separates a harmless biological artefact from a true treatment failure. Confirming which is which is now becoming possible in the clinic.

Could this change whether people with HIV can get surgery or join cure trials?

Potentially, yes. A confirmed detectable reading has sometimes excluded people from procedures like joint replacements or organ transplants, and from clinical trials with strict viral-load criteria. If a clinician can show the detectable virus is defective and noninfectious, those barriers may fall away. The cost-effective new test was designed partly with that use in mind.

What’s stopping this test from being used everywhere tomorrow?

Mainly genetic diversity. The assay was validated almost entirely on the HIV subtype common in North America and Europe, and it needs further testing against the wider range of strains circulating across Africa and Asia before global use. The researchers also want to push its sensitivity down toward single-copy detection. Both look achievable, but neither is done yet.

That wait may be over. A team at Chiba University in Japan has assembled bisleuconothine A in the lab atom by atom, the first time the compound has ever been built from scratch.

To understand why this took so long, you have to look at what the molecule actually is. Bisleuconothine A belongs to a sprawling family called monoterpenoid indole alkaloids, or MIAs, and it is one of the awkward oligomeric ones, meaning it is stitched together from more than one alkaloid unit into a big, lumpy three-dimensional shape. That bulk is precisely what makes it interesting to drug developers. Conventional small-molecule drugs tend to be flat and tidy, and they are not much good at jamming themselves into the interfaces where two proteins meet, but a large, contorted molecule like this one might just be able to wedge itself in there and break the contact.

Disrupting those protein-protein interactions is a long-standing dream in oncology. The trouble has always been getting hold of enough of the molecules that can do it.

Plants assemble these things effortlessly, of course, through enzymes refined over millions of years. Chemists, working without that machinery, face a structure riddled with interlocking rings and a fistful of stereocenters, the points where the molecule’s atoms have to be arranged in one specific handedness and no other. Get a single one wrong and the biological activity can simply vanish. So drug research on oligomeric MIAs has limped along, starved of material.

One Building Block to Rule Them All

The Chiba group, led by Hayato Ishikawa, went after the problem sideways. Rather than grinding out a bespoke route for each molecule, they built a single versatile fragment first and worked outward from there.

The fragment in question is a chiral 3-ethylpiperidine scaffold, a small ring system that crops up again and again across the MIA family, and the team coaxed it into existence using what is known as organocatalysis: chemistry driven by small organic molecules rather than the metal catalysts that dominate so much of synthesis. The reaction runs as a cascade, several transformations firing in sequence in a single pot. A chiral amine catalyst nudges a Michael addition along, locking in the correct handedness from the off, and a couple of further steps (a cyclization, then an acetalization) tidy the piece into a pure, reusable intermediate. Crucially they needed only a whisker of catalyst to do it. From that one common building block they then fashioned two different alkaloid fragments and stitched them together with a coupling reaction deliberately designed to mimic how a plant might do the join itself.

That biomimetic step is the elegant bit. It delivered bisleuconothine A in 20 steps.

And then it kept going. With one additional step the same approach produced a second, even more elaborate molecule called bousigonine B, a trimeric MIA built from three alkaloid units rather than two, marking the first time anyone has completed the total synthesis of a trimeric MIA at all. The work, published in Angewandte Chemie International Edition on 23 May, also quietly corrected the record: the team found that the accepted absolute stereochemistry of bousigonine B was wrong, and their synthesis revises it.

From Bench to Bedside, Eventually

None of this makes a cancer drug tomorrow. A total synthesis is a proof that the molecule can be made and made cleanly, not a finished therapy, and there is a long road of biological testing between a flask of pure compound and anything a patient might receive. Still, having a reliable supply changes what is possible to even attempt.

“The present method for total chemical synthesis is expected to facilitate the development of new pharmaceutical agents. In particular, bisleuconothine A has exhibited potent anticancer activity, highlighting its potential as a lead compound for anticancer drug development,” says Ishikawa.

The bigger prize, arguably, is not either molecule on its own but the strategy that yielded both. Because so many alkaloids share that 3-ethylpiperidine core, a single well-behaved intermediate could in principle open the door to a whole shelf of related natural products, the sort of compounds that have tantalized chemists precisely because they were too fiddly to make in any quantity. The team is already pressing in that direction. “Current efforts are directed toward the collective total synthesis of additional MIAs based on this newly established methodology, as well as subsequent biological evaluation for drug-discovery applications,” says Ishikawa. Which is a careful way of saying they intend to make a lot more of these molecules, and then find out what they do.

DOI / Source: 10.1002/anie.6698305 (Angewandte Chemie International Edition)

Frequently Asked Questions

Why does building a molecule in the lab matter if it already exists in a plant?

Plants make these compounds in tiny amounts and only under their own enzymatic control, so extracting useful quantities from bark is slow and unreliable. A lab route gives chemists a dependable, scalable supply they can tweak and study. That is usually the difference between a curiosity and a viable drug candidate.

How does a big, awkward molecule fight cancer differently from an ordinary drug?

Most small-molecule drugs are compact and struggle to interfere with the broad contact points where two proteins bind. A large, three-dimensional alkaloid like bisleuconothine A may be able to wedge into those interfaces and break the interaction. Disrupting protein-protein contacts is a target conventional drugs often cannot reach.

Is a cancer treatment now close at hand?

Not yet. A total synthesis proves the molecule can be made cleanly, but extensive biological testing still stands between a pure compound and an approved therapy. What changes is that researchers finally have enough material to do that testing properly.

What makes the trimeric molecule, bousigonine B, a notable first?

Bousigonine B is built from three linked alkaloid units, and no one had ever completed the total synthesis of a trimeric MIA before. The Chiba team also discovered that the compound’s accepted three-dimensional handedness was recorded incorrectly, and their work revises it.

That uncomfortable scenario was, until very recently, trivially easy to pull off against most phones in the United States. A team of computer scientists at the University of California, San Diego spent months pulling apart the plumbing that carries text messages, and what they found was a gap wide enough to walk a convincing impersonation straight through.

The trouble starts with a feature almost nobody remembers asking for. Back in the early 2000s, carriers wanted to popularise texting, so they wired up gateways that let you send a text by emailing it. Email an address like a phone number at the carrier’s domain, and the message pops out the other end as an SMS. Convenient, sure. But email and text are different languages, with different rules about who sent what, and somebody has to translate between them.

That somebody is the gateway, and translation is where things get lost.

“Email and text messaging weren’t designed to work together,” says Stefan Savage, a professor in UC San Diego’s Department of Computer Science and Engineering and one of the paper’s senior authors. He reaches for an image to make the awkwardness concrete: “It’s a little bit like reading postcards to someone over the phone and needing to figure out where the sender and recipient information and the message itself are.”

And every carrier figures it out slightly differently. The team probed the gateways of the big U.S. networks, Verizon, T-Mobile, Google Fi, AT&T, and a clutch of smaller operators, treating each one as a black box and feeding it malformed email after malformed email to see what came out. Email systems do have anti-spoofing defences with names like SPF, DKIM and DMARC, and the gateways all claimed to support them. The problem was the seams. Leave one header empty here, slip a stray character there, and a message that should have been rejected sailed through wearing whatever sender name the attacker fancied.

How a Stray Character Becomes a Phone Number

The really clever part happens once that email-turned-text reaches the handset. Phones try to be helpful: they check the sender against your contacts and show you a friendly name instead of a string of digits. The researchers discovered that a carefully crafted email address could hijack that lookup. On an iPhone, an address beginning with a phone number followed by the characters “=?” gets chopped at exactly the wrong spot, and the bit left over is read as a genuine phone number. Android had its own version of the bug, where Google Messages saw an all-numeric email address, decided it must really be a number, and quietly stripped out the @ and the dot until it became one.

So an attacker does not just spoof some random email. They can make your phone believe a message came from a specific number, a five-digit short code, or even a plain word like the name of a bank.

What makes this genuinely nasty, rather than merely clever, is what phones do next. To keep your conversations tidy, messaging apps bundle everything from one contact into a single thread, whether it arrived by SMS, iMessage or anything else. Apple’s app is especially eager about this, the researchers found, merging messages across phone numbers and email addresses into one continuous conversation without flagging which channel each one came in on. Spoof the right identity and your forged message does not start a suspicious new thread. It drops into the middle of a real one. There is a small caveat for the attacker, mind you: they generally cannot see the replies, since those go to the real contact.

“There are no standards for converting emails to texts and that opens the door to all sorts of vulnerabilities,” says Sumanth Rao, the paper’s first author and a computer science PhD student at the Jacobs School of Engineering.

The technical requirements for an attacker are, frankly, depressingly modest. You need a computer that can send email, some fiddly off-the-shelf software, a domain of your own, and the victim’s phone number, which is hardly a state secret. From the number you can usually look up the carrier, and from the carrier you can look up the gateway, because the carriers publish the addresses themselves. The same researchers also showed how to dress a forged message up as a “verified” business, complete with a recognisable logo, and how to fake an entire group chat in which the attacker plays every part except the victim.

The Long Tail of a Twelve-Year-Old Bug

Some of this had been sitting in plain sight for an alarmingly long time. The iPhone parsing quirk appears in Apple’s libraries going back to at least 2012, and the Android one to around 2016. These were not freshly minted holes; they were old assumptions nobody had thought to stress-test.

The whole edifice, the researchers argue, rests on a quiet bit of faith that none of us agreed to. We assume a text is what it says it is. “People don’t realize that there’s no guarantee that text messages have integrity,” says Savage. “You can’t count on authenticity.”

Here is the better news. Before publishing, the team disclosed everything to the affected companies, and the response was unusually brisk. T-Mobile patched its gateways within a day of being told; Verizon within five. Google fixed the flaw in Google Messages and Apple fixed the iPhone parsing bug, assigning it a formal vulnerability identifier in the process. Verizon is going further and plans to switch off the ability to send texts by email altogether by the end of March 2027, a path AT&T had already taken. The industry’s standards body, the GSMA, is updating its security guidance so carriers elsewhere can tighten the same loose joints.

So the front door has been bolted, at least in the US. What the work really exposes, though, is less a single bug than a habit of building. Whenever two old systems that were never meant to talk are bolted together for convenience, the gaps in the translation become someone’s opportunity, and those gaps tend to lurk for years before anyone goes looking. The next one is probably already out there, waiting in the seam between two services nobody thought to question.

The research, “Lost in Translation: Text Message Spoofing via Email,” received a Distinguished Paper Award at the 47th IEEE Symposium on Security and Privacy.

Frequently Asked Questions

Could someone really fake a text from my bank without hacking anything?

Yes, and that was the unsettling core of this research. By emailing a carrier’s text gateway with a few deliberately malformed details, an attacker could make a phone display their message as coming from a trusted name or number, no account breach required. The major US carriers and both Apple and Google have since patched the specific flaws, but the technique worked against ordinary phones for years.

Why was a forged text able to slip into an existing conversation?

Messaging apps group everything from one contact into a single thread to keep things tidy, and they tend to trust the sender label without verifying it. Apple’s app was the most aggressive, merging messages across email and phone numbers into one conversation, so a spoofed message could appear mid-thread rather than starting a suspicious new one. That bundling is convenient, but it quietly assumes every sender is who they claim to be.

Is my phone safe now?

For the specific attacks in this study, largely yes, provided your phone is updated, since Apple, Google and the major carriers have deployed fixes. Verizon is even planning to retire email-to-text entirely by early 2027. The deeper lesson is harder to patch: similar translation gaps may exist wherever two incompatible systems have been stitched together.

How hard would this have been to actually carry out?

Surprisingly easy by the standards of serious attacks. It needed only a computer able to send email, some common software, a domain, and your phone number, which is rarely hard to find. That low barrier is exactly why the researchers treated it as urgent rather than theoretical.

They gathered in Bloomington, Indiana, in February last year, convened by Indiana University’s School of Public Health under a deliberately combative banner: a critical assessment of postulated propositions. The idea was to take 11 widely held beliefs about dietary protein and, in the organizers’ phrase, pressure-test them.

What makes the exercise interesting is the verdict. The experts rated each proposition on a scale running from “existing evidence strongly supports” all the way down to “evidence seems sufficient to rule this out,” and for the majority of the 11, they landed somewhere in the uneasy middle. Plausible, maybe. Supported by a bit of preliminary data. But not actually nailed down, not by the standards scientists would demand in almost any other field. “Protein science has advanced significantly, but despite thousands of published studies, in some instances there is still a lack of publicly available quality data,” says Mitch Kanter, the workshop’s first author and a primary organizer.

Take the claim you’ve probably heard most often: that protein keeps you full, that it’s the most satiating of the three macronutrients and so helps you eat less overall. It feels true. It’s the entire pitch behind half the products on those shelves.

And it’s the one proposition the panel came closest to rejecting outright. The reviewers concluded that acceptance couldn’t be completely ruled out, but that the evidence actually points toward implausibility. The trouble runs deep. Appetite itself is fiendishly hard to measure, leaning on people’s self-reported hunger ratings, and those ratings turn out to be poor predictors of what anyone actually eats. One review of more than 460 studies found no link between appetite sensations and food intake in roughly half of them. So even if protein does nudge fullness upward a little, and some trials suggest it does, that nudge may simply vanish in the messy reality of a normal day’s eating.

The Thin Foundation

This is the uncomfortable thread running through the whole report. We have built sweeping public guidance on a surprisingly thin stack of evidence.

Consider one number that ought to give anyone pause. When a US government agency went looking for high-quality studies linking protein intake to bone disease, kidney disease, and muscle loss, it screened more than 11,000 papers covering a quarter-century of research. After filtering for rigour, the number that survived to the final analysis was 13. Thirteen. Just one of those concerned kidney health. For comparison, a single meta-analysis of cholesterol-lowering statins drew on data from over 50,000 people; the average protein feeding trial, across six pooled analyses, enrolled 49. There simply hasn’t been the money, or perhaps the will, to run the long, large, expensive trials that would settle these questions properly.

There’s a subtler problem too, and it goes to the heart of how protein gets studied. A great deal of the muscle research rests on something called muscle protein synthesis, measured over a few hours after a meal using tracer techniques. It’s an elegant tool. But synthesis rates are a marker for tissue turnover, not a reliable stand-in for whether you actually end up with more muscle months down the line, and the report is blunt that the two have been conflated for years. Much of what gets sold as settled muscle science is really a snapshot lasting less than a day.

What Actually Survived

Not everything got the skeptical treatment, mind you. A few propositions came through looking robust. The idea of protein leverage, that we keep eating until we hit a protein target and overshoot on calories when our food is protein-dilute, drew strong support, and may help explain why ultra-processed diets tip people toward overeating. The notion that individual amino acids do distinct jobs in the body held up well, as did the genuinely surprising finding, mostly from animals, that restricting certain amino acids can extend lifespan. And the old worry that high protein wrecks your kidneys or leaches calcium from your bones? For healthy people, the panel found that one essentially ruled out.

Older adults, meanwhile, probably need more protein than the official guideline of 0.8 grams per kilogram of body weight a day, possibly half as much again or more, to hold on to muscle as they age. Which means the “more protein” message isn’t wrong for everyone. It’s just been flattened into a slogan that ignores who’s actually asking.

One thing worth keeping in plain view: the workshop and the resulting paper were funded by a long roster of food-industry players, the National Pork Board, dairy and beef groups, egg and soy interests, big manufacturers. The journal’s own editor-in-chief was an author, though he recused himself from reviewing it. None of that makes the findings wrong, and the report is unusually candid about its own funding. But it does lend a certain piquancy to an industry-backed panel concluding, in effect, that the protein hype has outrun the data. The pork board’s own nutrition director frames the takeaway as shifting “from ‘more protein’ to ‘better protein’,” which is a tidier marketing line than “we’re not actually sure.”

What the report really leaves you with is less a set of answers than a long list of honest questions, and a quiet rebuke to the certainty of the marketing. We know protein is essential. We know some people, the elderly especially, likely need more of it. Beyond that, the picture blurs fast. The next time a label promises that a scoop of something will keep you fuller, build you bigger, or extend your years, it’s worth remembering that the people who study this for a living are still, in most respects, working it out.

The full review is published in Critical Reviews in Food Science and Nutrition.

Frequently Asked Questions

Is it true that protein keeps you fuller than carbs or fat?

It’s far less certain than the marketing implies. The expert panel rated this proposition as leaning toward implausible, partly because appetite is notoriously hard to measure and self-reported fullness is a weak predictor of how much people actually eat. Protein may modestly increase fullness in some studies, but that effect often fails to translate into eating less over a real day.

How much protein do older adults actually need?

Probably more than the standard recommendation of 0.8 grams per kilogram of body weight per day. The evidence reviewed suggests intakes of around 1.2 to 1.6 grams per kilogram, or even higher, may help preserve muscle mass and function with age. Current dietary guidelines in most countries don’t yet differentiate by age this way.

Does a high-protein diet damage your kidneys?

For people with healthy kidneys, the review found no evidence of harm at intakes up to roughly 1.5 grams per kilogram a day or 20 percent of energy. The long-standing fear traces back to studies in people who already had kidney disease, which is a different situation. Whether decades of very high intake matters is still genuinely unknown.

Why is the protein evidence base so weak if there are thousands of studies?

Quantity isn’t quality. Most protein trials are small, short, and rely on surrogate markers like muscle protein synthesis rather than long-term outcomes like actual muscle gain or disease risk. When one major review screened over 11,000 papers, only 13 met the bar for high-quality evidence, which tells you how thin the rigorous foundation really is.

Should I stop trusting high-protein products?

Not necessarily stop, but read them with a healthier skepticism. Protein is essential and some groups clearly benefit from more of it, yet many specific claims on packaging run well ahead of what the science has actually demonstrated. The honest position, even among researchers funded by the food industry, is that a lot remains unproven.

The finding comes from a team at the University of Arizona, working with colleagues at the University of Southern California, who went looking for the fingerprints of poor sleep in brain scans taken years after people first answered questions about their nights. What they found complicates the comforting idea that sleep is simply good or bad, and quite a lot in between.

The brain marker in question is called a white matter hyperintensity, which is a mouthful for something fairly simple to picture: small patches of damage in the brain’s wiring that show up as bright spots on an MRI scan. They accumulate as we age, more in some people than others, and a heavier load of them is linked to a higher risk of dementia, including Alzheimer’s disease. The bright spots are thought to reflect the slow deterioration of the tiny blood vessels that keep brain tissue supplied. In other words, they are a window onto how well the brain’s plumbing is holding up.

So the question the team set themselves was deceptively narrow. Which sleep habits, if any, leave their mark on that plumbing?

To answer it they turned to the UK Biobank, a vast trove of health data drawn from half a million British volunteers. Participants had filled in a questionnaire between 2006 and 2010 about five aspects of their sleep: how long they slept, whether they napped, whether they struggled to drop off or stay asleep, whether they snored, and whether they dozed off when they didn’t mean to. Roughly nine years later, more than 23,000 of them came back for brain imaging.

“Sleep is a universal but complex behavior, and there is still much to learn about how different aspects of sleep relate to brain health,” says Madeline Ally, the study’s lead author and a graduate researcher in the university’s psychology department. That complexity is the whole point. Sleep is too often boiled down to a single score, good or bad, and the texture gets lost.

Sorting Signal From Noise

At first, every one of the five habits looked guilty. Run the numbers with only the basic adjustments for age, sex, education and the like, and all five, snoring included, lined up with more white matter damage. But poor sleep travels in bad company. People who sleep badly are also more likely to have high blood pressure, to smoke, to carry extra weight, to move less, and any of those could be doing the damage instead. So the researchers stripped those factors out, one cardiovascular risk at a time, to see what was left standing on its own.

Two of the five suspects fell away. Snoring, despite looking like the strongest culprit early on, faded once blood vessel health was taken into account, as did daytime dozing. Three habits survived the cull: sleeping outside the recommended seven-to-nine-hour window, frequent napping, and persistent sleeplessness, the broken nights and the difficulty drifting off. Each appeared to leave its own distinct mark, independent of the others and of the usual cardiovascular suspects. And napping, oddly enough, carried the heaviest weight of all.

That sits awkwardly against everything else we hear about naps, which is rather the point. Short naps have been linked, in other work, to sharper thinking and better alertness. The catch is that this questionnaire never asked how long anyone’s naps actually were, or when they happened. A restorative twenty-minute doze and a daily two-hour collapse on the sofa got counted as the same thing. Gene Alexander, the senior author and a psychology professor at Arizona, points to that gap as the obvious next thing to chase down: whether brief, occasional naps behave differently in the brain over time than long, habitual ones.

The sleep-duration story had its own twist. When the team looked harder, the damage was concentrated among the short sleepers, those getting under seven hours, rather than the long ones. “Our findings suggest that having too little sleep may lead to greater white matter lesion volumes in the brain as we age,” says Alexander. “We didn’t see greater white matter impacts in people who reported longer sleep durations, but this needs to be followed up in cohorts with more long sleepers.” Honest enough; there simply weren’t many long sleepers in the sample to draw firm conclusions from.

A Risk You Can Actually Change

It’s worth keeping the limits in view. This is an observational study, which means it can spot a pattern but cannot prove that bad sleep is causing the damage rather than the other way round, or that some third thing drives both. The sleep data was self-reported, never the most reliable sort. And the UK Biobank skews overwhelmingly white and was screened down to healthy adults, so whether the same holds across more varied populations remains an open question.

Still, there’s a reason the researchers keep circling back to these three habits in particular. Unlike your age or your genes, they are things you can do something about. “Sleep is one of those potentially modifiable risk factors,” says Alexander. “If we can improve the quality of our sleep, it may help reduce the impacts of brain aging and maybe even lower the risk for dementias like Alzheimer’s disease.” Whether tidying up your sleep can actually slow the bright spots from spreading is the thing nobody has tested yet, and it’s the experiment that would matter most.

Source: Ally et al., Alzheimer’s & Dementia (2026), DOI 10.1002/alz.71457

Frequently Asked Questions

Is napping actually bad for your brain?

Not necessarily. This study found that frequent napping was the habit most strongly tied to a marker of brain aging, but the questionnaire never recorded how long the naps were or when they happened, so a short refreshing doze and a long daily one were lumped together. Other research has linked brief naps to better alertness and thinking, so the likely answer is that duration and regularity matter a great deal, and untangling that is the next job.

Why does sleeping too little seem to damage the brain?

The damage shows up as white matter hyperintensities, small bright patches on a brain scan that reflect ailing blood vessels in the brain’s wiring. During healthy sleep, heart rate and blood pressure normally drop, which is thought to protect those vessels over the long run, so consistently short nights may chip away at that protection. The link held up even after accounting for high blood pressure, smoking and weight, suggesting sleep plays its own role beyond the usual cardiovascular suspects.

Does getting more sleep protect you, then?

The study could not say. The brain damage clustered among short sleepers getting under seven hours, while long sleepers showed no clear effect, though there were too few of them to draw firm conclusions. And because the work only observed people rather than testing changes, it cannot prove that fixing your sleep would slow the damage, which is exactly the experiment researchers say still needs doing.

Should I be worried if I sleep badly?

A single rough patch is not what this is about; the study looked at long-standing habits across tens of thousands of people. The encouraging part is that all three flagged behaviors, short sleep, frequent napping and persistent sleeplessness, are things you can potentially change, unlike your age or genes. That makes sleep a rare modifiable lever on brain aging, even if the precise payoff has yet to be measured.

What if you could keep that cell’s batteries charged? Researchers at UCLA think they have found a way, and it involves something startlingly mundane: creatine, the same white powder that athletes and bodybuilders have been scooping into their shakes for decades.

The study, published in April in iScience, builds on earlier work from the same lab showing creatine powers the killer T cells themselves. This time the team went looking one step upstream, at the cells that give those killers their orders. They started by reading which metabolic genes were busiest in dendritic cells that had pushed their way into mouse tumours. One stood out. The gene encoding the creatine transporter, the little protein that hauls creatine into a cell, was running far hotter inside tumours than in healthy tissue. The cells, it seemed, were already reaching for the stuff.

So the team took the transporter away. Dendritic cells engineered to lack it survived poorly, activated weakly, and made a feeble case to the T cells they were supposed to recruit.

Then they did the opposite. They gave mice with melanoma a daily dose of creatine, and the tumours grew more slowly. The treated animals had more of the potent antigen-presenting cells crowding into their tumours, and those cells were buzzing, churning out chemical signals that pull yet more immune reinforcements into the fight.

A battery for the immune system

Why should a muscle-building supplement do any of this? The answer is that creatine is, in a sense, a rechargeable battery. Inside the cell it shuttles high-energy phosphate back and forth, soaking up spare energy and releasing it on demand, keeping levels of ATP, the molecular currency that powers nearly everything a cell does, steady even when the surroundings turn lean. The metabolomics bore this out: creatine-fed dendritic cells held onto more ATP and kept their inflammatory machinery humming. Take the transporter away and that energy buffer collapses, and with it the signalling that tells a dendritic cell to switch on.

“Immunotherapy has shown remarkable promise, but it only works for a subset of patients,” says Lili Yang, the study’s senior author and a professor of microbiology, immunology and molecular genetics at UCLA. Most approved immunotherapies aim squarely at the killer T cells at the end of the chain, yet only roughly 20 to 40 per cent of patients respond. The thinking here is to support the whole apparatus instead, the cells that spot the threat and set everything in motion, not just the soldiers who finish the job.

“Understanding how to metabolically support dendritic cells is about supporting the entire anti-tumor response, not just the killer T cells at the end of it,” says Elliot Kang, a co-first author who worked on the study as an undergraduate in Yang’s lab.

From the gym to the clinic, maybe

The effect was not confined to mice. When the researchers treated human dendritic cells, the kind grown from blood and used to build dendritic cell cancer vaccines, creatine sharpened them up too, improving their ability to rouse human T cells against a cancer target. That hints at two rather different uses. “The potential we see here is that creatine could be used in two complementary ways: as a supplement to enhance the immune response of patients already receiving immunotherapy, and as a tool to improve the quality of dendritic cell-based vaccines before they’re administered,” says James Elsten-Brown, a co-first author and graduate student in the lab.

A note of caution, and the researchers are firm about it. This was done in cells and mice, not people. Nobody should be reading it as a reason to start dosing themselves mid-treatment, and anyone on cancer therapy should talk to their doctor before adding any supplement at all. There is a further wrinkle worth keeping in view: creatine is a ubiquitous energy metabolite, and a handful of studies suggest tumour cells can hijack it too, using it to fuel their own spread. The full picture, across immune cells and malignant ones alike, is not yet settled.

Still, creatine is cheap, and after decades of use in gym bags its safety profile is about as well documented as any supplement going. The team now hopes to work with physicians on trials that would test whether it actually helps patients on immunotherapy. The humble scoop of powder has had a long career building biceps. Its second act, it seems, might be fought somewhere far less visible, in the starved interior of a tumour, one tired immune cell at a time.

Frequently Asked Questions

How does creatine actually help fight cancer?

It does not attack tumours directly. Instead it acts as an energy buffer inside dendritic cells, the immune cells that identify a tumour and direct killer T cells to attack it. By keeping those cells supplied with ATP even in the nutrient-starved environment of a tumour, creatine helps them stay active and raise a stronger alarm, which in turn mounts a more aggressive T cell response.

Should people with cancer start taking creatine?

Not on the strength of this study. The work was done in cells and mice, not patients, and the researchers explicitly warn against drawing any medical conclusions from it. Anyone undergoing cancer treatment should consult their doctor before adding creatine or any other supplement, particularly since some research suggests tumour cells can also exploit creatine.

Why does immunotherapy only work for some patients?

Most approved immunotherapies target killer T cells directly, but only about 20 to 40 per cent of patients respond. One reason may be that the supporting cells that activate those T cells, including dendritic cells, are themselves worn down inside tumours. Energising that wider infrastructure, rather than the T cells alone, is the strategy this research points toward.

Could creatine improve cancer vaccines?

Possibly. Dendritic cell vaccines are made by growing a patient’s own dendritic cells in the lab, and these are the most common cells used in such platforms. Adding creatine during manufacturing boosted the activating power of human dendritic cells in the study, suggesting it might be used to make more potent vaccines, though this remains to be tested in clinical trials. You can read the full study in iScience.

For most of the people whose teeth they read, the answer was no. Not really, not on purpose, not often enough to leave a mark.

This matters more than it might seem, because right now there is a small industry trying to convince Western consumers that insects are the protein of the future. The Food and Agriculture Organization has been making the case for years: 1,611 edible species cataloged, hundreds of millions of people already eating them happily around the tropics, a far lighter footprint than cattle or pigs. And yet most Europeans recoil. The usual explanation is cultural, something to do with medieval Christianity reclassifying locusts as biblical plagues, or simple unfamiliarity. The new work, published in Science Advances, suggests the aversion runs a good deal deeper than that.

To get at it, Librado’s group did two very different things at once. First the tartar, screened against a custom library of more than ten thousand insect mitochondrial genomes. Then the human genome itself, scanned for the genes that let us digest chitin, the tough stuff that makes up an insect’s shell.

Both lines of evidence pointed the same way. In the dental calculus of ancient northern Eurasians, insect DNA was scarce, and what little turned up looked accidental: midges whose larvae live in lake sediment, suggesting a mouthful of pondwater rather than a meal; pests associated with damp grain stores. The team compared these humans against a benchmark of 96 great apes and found something telling. Ancient Europeans carried about as much insect DNA in their tartar as chimpanzees from the rainforests of Uganda, animals for whom insects make up less than four per cent of the diet because fruit is simply easier.

“The scarce presence of insects in the diet of northern Eurasians suggests that the absence of entomophagy is not solely due to recent cultural factors, but also to a long ecological and evolutionary history,” says Librado, who led the study.

The Neanderthals Were Different

Here the story takes a turn, because not everyone in ancient Europe was so squeamish. The 18 Neanderthals in the sample carried insect DNA at levels comparable to western chimpanzees, the savanna-edge apes that genuinely rely on termites to get through lean seasons. And the dominant signal in Neanderthal tartar was flies, mosquitoes especially, which is a strange thing to find in someone’s mouth until you consider what it might mean. The researchers connect it to a recent and slightly grim hypothesis: that Neanderthals routinely ate the carcasses of their kills after the maggots had moved in, and that they may have stored those carcasses in ponds and marshes, exactly the boggy places where mosquitoes lay their eggs. One Neanderthal from a Belgian cave, known as Spy, yielded DNA from a moth fly whose larvae feed on late-stage rotting flesh. It was the single clearest molecular fingerprint of maggot-eating in the whole study, supported by two reads out of 13.5 million tested. Two reads. The forensic standard here is brutal.

And the genetics agreed with the tartar. Both the Neanderthals and the lone Denisovan in the dataset carried versions of the chitin-digesting genes tuned for breaking down exoskeletons, the very versions that most modern Europeans lack.

It is worth dwelling for a moment on how easily this kind of work goes wrong, because the team is admirably blunt about it. Dental calculus is not the sealed vault everyone assumed. One 42,800-year-old Neanderthal turned up DNA from the harlequin ladybird, a species that did not set foot in Europe until the 1980s, when it was introduced for pest control. The skeleton, in other words, had been quietly contaminated by a beetle that arrived forty millennia after its owner died. If a modern ladybird can infiltrate a clean Pleistocene skeleton, you have to wonder what else has slipped in, and the researchers wonder it openly.

A Map Written in the Gut

The genome scan is where the argument hardens. Two genes, CHIA and CTBS, encode enzymes that work in the stomach to break down chitin, and when the team looked at how their variants are distributed across the planet, they found one of the strongest geographic gradients anywhere in the human genome, ranking in the top fractions of a per cent. The pattern tracks latitude almost eerily: the closer a population’s ancestral home to the tropics, the better its genetic equipment for digesting insects. The reason, the team argues, is sheer arithmetic. “Large quantities of insects need to be ingested to compensate for the high caloric expenditure involved in their collection. In the tropics, there is a greater availability of social insects, such as termites and locusts: their biomass and diversity allow for sustainable exploitation throughout the year, which even contributes to pest control,” explains Manuel Piñero, the study’s first author. Move north, where insects are seasonal and scattered, and the calories you burn hunting them stop being worth it.

What makes the gradient so striking is its age. By reaching into 1,663 ancient genomes, the team showed that this latitudinal cline was already in place at the dawn of farming, some 9,000 years ago, and that it has held steady ever since, through the vast migrations that otherwise reshuffled Europe’s ancestry. Relict hunter-gatherer groups such as Japan’s Jomon already carried the low-digestibility variants, which means the predisposition predates agriculture rather than following from it.

So when a modern European wrinkles their nose at a plate of crickets, they may be expressing something older than disgust, older than Christianity, older than the plough. According to Librado, the trail leads back to ecology. “Beyond cultural or religious factors, our results suggest that the reduced availability of insects in non-tropical areas may have been a key factor in the abandonment of entomophagy, leading to a reduced capacity to digest insect exoskeletons.” The taste went away, the gene followed, and the culture eventually wrote a story to explain a feeling whose real roots lay in the landscape.

None of which dooms the cricket-flour cookie. The whole problem is the exoskeleton, the indigestible chitin, and modern processing can simply strip it out, leaving the protein behind, which is rather the point of farming insects at industrial scale in the first place. The aversion, it turns out, may be a few thousand years of ecology talking. It just might not get the last word.

Read the study in Science Advances

Frequently Asked Questions

Why would Europeans struggle to digest insects when people elsewhere eat them every day?

The shells of insects are made of chitin, and breaking it down depends on stomach enzymes encoded by genes that vary by geography. Populations whose ancestors lived near the tropics, where insects are abundant year-round, tend to carry versions tuned for efficient chitin digestion, while northern populations carry variants that do this poorly. The new genomic work suggests this difference has been baked into our DNA for at least 9,000 years.

How can anyone tell what someone ate 30,000 years ago?

The answer is dental tartar, technically called calculus, which hardens on teeth and traps fragments of DNA from food before fossilizing alongside the skeleton. By screening this calcified residue against a library of insect genomes, researchers can detect traces of what a person regularly consumed. The same method also catches accidental contamination, which is why the team applied punishingly strict standards before counting any signal as real.

Is it true that Neanderthals ate maggots?

The evidence points that way, though gently. Neanderthal tartar contained far more insect DNA than that of contemporary modern humans, dominated by flies and mosquitoes, which fits a hypothesis that they ate decomposing carcasses colonized by fly larvae and possibly stored those carcasses in water. One specimen even yielded DNA from a fly whose larvae feed on late-stage rotting flesh, the clearest sign yet of maggot-eating in the record.

Does this mean farmed insects will never catch on in the West?

Not necessarily. The digestive obstacle is the chitin in the exoskeleton, and modern food processing can separate that out, leaving behind the protein that makes insects nutritionally attractive in the first place. So the ancient aversion may shape our instincts without dictating whether insect-based foods eventually find a place on European plates.

On 1 June, the International Academy of Astronautics ratified a sweeping update to the protocols that would govern how scientists tell humanity it has company. The last version was set down in 2010, back when a tweet was a novelty and the phrase “deepfake” did not exist.

The document carries a dry title, the Declaration of Principles Concerning the Conduct of the Search for Extraterrestrial Intelligence, and it does something the public conversation about aliens almost never does: it slows everything down. Behind it sits Michael Garrett, who holds the Sir Bernard Lovell Chair of Astrophysics at the University of Manchester and chairs the IAA’s SETI committee. He led a revision that pulled in more than 350 researchers across several years. The trigger was not a signal. It was the world the signal would land in.

“The information environment we operate in today is vastly more complex than it was in 2010,” says Garrett. The worry is not the discovery itself but the few hours after it, when a half-confirmed rumour could outrun every careful caveat.

And the search has changed almost beyond recognition. When the first version of these principles was drawn up in 1989, SETI meant listening for a narrow-band radio whistle, the cosmic equivalent of a struck tuning fork. The new declaration takes a much wider view. Researchers now comb the whole electromagnetic spectrum: optical laser pulses, multi-messenger signals, and the faint infrared glow that a vast alien megastructure would leak as waste heat. A so-called technosignature, in the document’s language, might even be a physical artefact.

Check, Check Again, Then Ask Others to Check

At the core of all this is an old principle wearing new armour: extraordinary claims require extraordinary evidence. Under the revised rules, nothing goes public until a candidate has been authenticated by independent groups, on different instruments, working with different methods, the scientific equivalent of refusing to trust a single witness.

“We do not shout “alien” the moment we see a strange blip,” Garrett adds. “The scientific method demands we check, check again, and then ask others to check. Only when we have reached a consensus that a signal is credible do we bring it to the world.” It sounds almost too patient for the subject matter. That is rather the point.

What is genuinely new is the attention paid to the humans caught in the middle. The declaration acknowledges, in plain terms, that anyone connected to a credible detection could face harassment, doxxing, and a media crush of a kind few scientists are trained for. So it builds in protections. Researchers may decline to engage with the press, or with social media, without it counting against them professionally; their institutions are expected to shield them and to keep the science flowing in their stead. There is a clause for the awkward case too, the candidate that turns out to be a satellite or a microwave oven in the building next door. If a signal proves earthly after all, the rules say, admit it promptly and clearly. No quiet burials.

There are limits to what a declaration can actually do. It is, in the end, a voluntary code, not a law; a rogue claimant with a big enough following could ignore every line of it. The committee seems aware that its real power is reputational, a shared standard that lets the serious distinguish themselves from the noise.

Nobody Gets to Answer for the Species

Then comes the part that has gripped people since long before anyone could detect anything. If we hear from them, do we answer? On this the declaration is immovable, and deliberately so. No reply is to be sent on anyone’s private authority. A response, it states, is a decision belonging to all of humanity, to be taken only after international consultation, specifically through the United Nations. The discoverers get to make the first announcement, but nobody gets to speak back for the species. A confirmed detection would be reported in full to the public, the scientific community, and the UN Secretary General, backed by a peer-reviewed report and the raw data, archived in at least two repositories on different parts of the planet.

To keep watch over the longer aftermath, the committee will stand up a permanent Post-Detection Sub-Committee, stocked not only with astronomers but with ethicists, lawyers, social scientists and people who study how risk is communicated. A formal presentation to the wider scientific world is scheduled for the International Astronautical Congress in Türkiye later this year. The rules are written, lodged, and waiting for an event that may never arrive, or may arrive tomorrow.

What the declaration really is, underneath the careful procedure, is a rehearsal. Humanity practising its lines for the most consequential phone call it could ever receive, in the hope that if the moment comes, we manage to act less like a crowd and more like a civilisation.

Full document: IAA Declaration of Principles Concerning the Conduct of the Search for Extraterrestrial Intelligence (2026 Update)

Frequently Asked Questions

Why do SETI scientists need rules for announcing a discovery at all?

Because the gap between spotting an odd signal and confirming what it is can stretch for weeks or months, and that gap is exactly where rumours, hoaxes and panic can take hold. The 2026 protocols are designed to keep verified facts separate from viral speculation during that vulnerable window. The thinking is that how a discovery is handled may matter almost as much as the discovery itself.

If aliens were detected, would scientists tell the public right away?

Not immediately, and not until a candidate signal has been independently checked by more than one organization using different instruments. There is no obligation to announce anything while verification is still under way, though scientists are expected to correct rumours and to confirm if a signal turns out to be a false alarm. Only a result that survives that scrutiny gets brought to the world.

Is it true that no one is allowed to reply to an alien signal?

Under the declaration, yes, at least not on any individual or single nation’s authority. Sending a response is treated as a decision for all of humanity, to be made only after international consultation through the United Nations. The rules cover replying to a confirmed detection; the separate question of broadcasting messages into space first is left to other agreements entirely.

What counts as a sign of alien technology these days?

Far more than the classic narrow-band radio signal. Researchers now look for optical laser pulses, unusual infrared heat that could leak from enormous engineered structures, and other anomalies across the electromagnetic spectrum, collectively called technosignatures. Notably, the protocols deliberately exclude UFO or UAP sightings in Earth’s atmosphere from their scope.

For decades this has been the quiet ceiling on a whole class of quantum machines. The setup, known as cavity quantum electrodynamics, or cavity QED, is the workhorse behind some of the most precise sensors ever built. But its symmetry has always held it back.

Now a team at the University of Chicago’s Pritzker School of Molecular Engineering has found a way to break the symmetry without breaking the apparatus. Their trick, described on 1 June in Physical Review X, uses tools already sitting in quantum labs around the world. No exotic new hardware. Just an extra magnetic field, or a second set of lasers, used to give different groups of atoms slightly different identities. The result is a recipe for highly entangled quantum states, some of which physicists had never thought to look for.

“The challenge has always been that these systems have too much symmetry. All the atoms are talking to light in the same way,” says Aashish Clerk, the molecular engineering professor who led the work. The team’s fix sounds almost too modest to matter.

Here is the idea. Each atom has a low-energy ground state and a higher-energy excited state, separated by a fixed gap. While every atom is driven by one common laser, the researchers nudge the excited-state energy of different groups up or down, pairing each group with a partner shifted by an equal and opposite amount. That small asymmetry gives the atoms distinct personalities while keeping enough structure for the maths to stay solvable. Change which atoms get which nudge, and the whole system settles into a different entangled state, all without touching a single physical component.

“You turn these lasers on and wait, and at some point the system stabilizes into an interesting, highly entangled quantum state,” says Anjun Chu, the postdoctoral researcher who is first author on the paper. “By simply adjusting the lasers, we can access kinds of entangled states that no one had thought about before.”

Turning leakage into a tool

What makes the approach genuinely clever is what it does with loss. In a cavity, light leaks out; photons escape, and that escape normally drains away the delicate quantum behavior you are trying to preserve. Most schemes for building entangled states fight this dissipation, or demand a long list of carefully engineered, independent loss channels to tame it. Clerk’s group did something closer to judo. They lean on a single, naturally occurring leakage process, the same collective decay that already happens in every cavity QED experiment, and use it, together with those paired energy shifts, to actively push the atoms toward the state they want. Turn on the drives, let the system bleed, and it relaxes not into mush but into a unique, pure, deeply entangled configuration. The escaping light, oddly, does the assembling.

The states that emerge have a peculiar internal bookkeeping. Atoms with equal and opposite energy shifts end up paired, their fates intertwined, and by reshuffling which atoms pair with which, the team can dial in entanglement of varying complexity. There is even a tidy mathematical analogy hiding in here: arranging the atoms into the right order turns out to mirror “bubble sort,” the schoolbook algorithm for sorting a list by swapping neighbors one pair at a time.

Sensing the difference between two places

The most immediate payoff is in quantum sensing. Entangled atoms can, in principle, detect impossibly small differences in a magnetic or gravitational field between two spots. The trouble is that entanglement is fragile, and the very states most sensitive to a signal tend to be the ones most easily wrecked by background noise. Engineers have long wanted a sensor that is both exquisitely sensitive and stubbornly robust, and the two demands usually pull in opposite directions. Clerk’s team showed that a version of their setup, using two clouds of atoms placed in two locations, can measure the gradient between the local fields while shrugging off noise that rattles both clouds equally.

“You’re able to do two things that are normally not compatible with one another: Use entanglement to build an exquisitely sensitive sensor but also have robustness to arbitrarily large amounts of noise,” says Clerk. “Normally, entanglement is very fragile. This approach has some amazing resilience.”

And you do not need fancy equipment to read the answer out. The states can be measured with standard Ramsey measurements, the bread-and-butter technique already used in atomic clocks and interferometers, which matters rather a lot if any of this is to leave the chalkboard. The same logic extends to four clouds of atoms, which the team showed could sense not just a gradient but the curvature of a field, the way it bends across space.

The reach goes beyond sensing, too. The same humble setup can be tuned to stabilize the AKLT state, a famous knot of many-body entanglement first dreamed up in the 1980s to describe exotic magnetic materials, and now prized as a possible resource for quantum computing. That such an elaborate state should fall out of lasers, mirrors and a bit of leakage is the sort of thing that makes theorists sit up.

For now it is all theory. The researchers are talking with experimental groups about putting the scheme to the test, and mapping out the wider zoo of states the method might reach. “The fact that such simple ingredients can generate such complex and useful quantum states gives us hope that even before we reach the dream of a general all-purpose quantum computer, we can already generate quantum states that let us do things we couldn’t do in a purely classical world,” says Clerk. The mirrors, it seems, have not finished surprising us.

Source: Chu et al., Physical Review X, 1 June 2026. DOI: 10.1103/qdh9-2pc7

Frequently Asked Questions

Why does breaking the symmetry of these atoms matter so much?

When every atom in a cavity responds to light identically, the system can only ever produce a narrow menu of entangled states, which caps what the technology can do. By giving groups of atoms slightly different energies, researchers unlock a far broader family of states from the same hardware. It is the difference between a piano with three keys and one with the full keyboard.

How can the same leaky cavity that ruins quantum states also build them?

Normally, light escaping a cavity drains the fragile quantum behavior physicists want to keep. The clever move here is to treat that single, natural leakage process not as an enemy but as a steering force that, combined with carefully tuned lasers, pushes the atoms into a precise entangled configuration and holds them there. The system settles into the target state rather than decaying away from it.

Could this actually improve real-world sensors?

That is the hope, and the design was built with practicality in mind. The states can be read out with Ramsey measurements, a standard technique already used in atomic clocks and interferometers, and they stay sensitive to a signal while ignoring noise that affects two locations equally. The work is still theoretical, so the next step is for an experimental group to put it to the test.

What is the AKLT state and why would anyone want one?

The AKLT state is a particular tangle of many-body entanglement first proposed in the 1980s to understand exotic magnetic materials, and it has since become a sought-after ingredient for certain approaches to quantum computing. Producing it usually takes considerable effort, so the fact that a simple arrangement of lasers and mirrors can stabilize it is what caught physicists’ attention here.

That, in a sentence, is the uncomfortable headline of the new State of the States report from the State of the Nation Project at Tulane University, released this week as the country edges up on its 250th birthday. It is not the story most of us expect about a polarized America.

The numbers behind it are enormous. Researchers pulled together more than three decades of data, from 1990 to 2024, across all 50 states and the District of Columbia, and ranked them on 31 separate measures, things like life satisfaction, trust in institutions, civil liberties, education, the environment, physical and mental health, inequality. More than 4,000 indicators in all. The point was to answer one deceptively simple question that the authors keep returning to: how are we actually doing?

The answer depends a great deal on where you stand. Minnesota came out on top with the strongest average ranking across every measure. Louisiana came in last. New England and the western Midwest clustered near the top; the three Southern census divisions sat at the bottom.

The Same Direction, Mostly Downhill

But the ranking, the part that lends itself to a tidy map, is almost the least interesting thing here. What unsettles is the direction of travel. For most of the measures showing a national decline, nearly every state is sliding the same way at once. No state is improving on adult depression. None on youth depression. None on life satisfaction, fatal overdoses, income inequality, or trust in the federal government. Eight measures, and on every one of them, the entire country is moving backward together.

“While all states are struggling with mental health, some states are getting hit harder than others,” says Anna Lembke, the Stanford psychiatrist who co-authored the report. That last clause matters. Because alongside the grim uniformity runs a second, quieter trend: on the measures that touch well-being most directly, states are not just declining, they are pulling apart from one another. Some are sinking fast. Others slowly. The gap widens.

There is a bright side, sort of, and it is worth saying plainly. Every single state has improved on two things over the long run: child mortality has fallen, and total real state income has climbed. Kids are surviving childhood at higher rates everywhere, and the economy, by the crude measure of dollars, keeps growing. Which makes the rest of the picture all the more puzzling.

The Money Doesn’t Buy What You’d Think

Here is the puzzle. If incomes are rising in every state, why is hardly anyone happier? The researchers went looking for the link and mostly couldn’t find it. Of 225 chances for a state to show improvement on a self-reported well-being measure, they counted just 12 cases where things actually got better. When they folded in overdoses and suicides, the picture darkened further still: 96 per cent of cases worsening, with a lone exception (Washington, DC, on suicide). And when they lined up income per capita against the personal measures, life satisfaction, adult depression, youth depression, the correlation simply wasn’t there. More money in the state, no measurable lift in how people felt about their lives.

Curiously, money did seem to track with something. Wealthier states tended to score better on social trust, the trust people place in institutions, in science, in one another. Not personal contentment, but the social fabric. The authors are careful not to overclaim which way the arrow points; well-functioning institutions might fuel a strong economy, or a strong economy might make institutions look like they’re working. Either reading is plausible. We don’t yet know.

What gives the project its odd authority is who built it. This is not a partisan broadside. The board behind the report draws on seven of the country’s leading think tanks from across the political spectrum, plus advisors to the last five presidents, Democrats and Republicans alike, Clinton through Trump. They are unpaid volunteers, and they agreed. “It’s not easy to capture how states are doing,” says Frederick Hess, a political scientist at the American Enterprise Institute and one of the authors. “This endeavor brought together a healthy mix of expertise and perspective, yet wound up with a remarkable degree of consensus as to what measures are most fundamental.” A group assembled precisely because its members disagree, finding common ground on the diagnosis if not the cure.

Douglas Harris, the Tulane economist who directs the project, frames the whole effort as a kind of shared mirror. At a time of polarization and pessimism, he argues, it matters to get a clear sense of how the country is really doing, and the surprise is that states red and blue mostly share the same struggles. The report stops there, deliberately. It prescribes no policy, names no villain. The authors say their job is only to establish a set of facts everyone can stand on before the arguing starts.

Maybe that’s the real provocation buried in 4,000 indicators. We spend a great deal of energy convinced that the country has split into two nations who want opposite things and live opposite lives. The data suggests something stranger and harder to fit on a yard sign: that we are, increasingly, the same anxious, distrustful, materially-richer-but-no-happier place, just arguing about it from different rooms. The question the authors leave hanging is the one worth sitting with. If the trouble is shared, what would it take to notice?

The full report and state-by-state findings are available at stateofnation.org.

Frequently Asked Questions

Is it true that red states and blue states are actually becoming more alike?

On most measures, yes. The Tulane analysis found states converging on 17 of the indicators it tracked and diverging on only 13, even as national political rhetoric grows more divided. The catch is that several of the measures pulling states apart are the ones tied most closely to well-being, which may be feeding the sense of division more than the underlying data warrants.

Why doesn’t rising income make people in wealthier states happier?

That is exactly the disconnect the report highlights, and the authors don’t claim to fully explain it. Across 225 opportunities for states to improve on self-reported well-being, only 12 actually did, despite incomes climbing nearly everywhere. Money did track with stronger social trust, but not with personal life satisfaction, which suggests whatever drives contentment runs deeper than the size of a paycheck.

How can a bipartisan group agree on something this politically charged?

The board was deliberately built from people who disagree, drawing on seven think tanks across the political spectrum and advisors to the last five presidents of both parties. They didn’t have to agree on solutions, only on which measures matter and what the data shows. The fact that such a group reached consensus on the diagnosis is, the authors argue, a small sign the country is less divided than it feels.

What’s stopping the report from telling states how to fix any of this?

That’s by design rather than oversight. The authors set out to establish a shared set of facts first, on the logic that you can’t agree on how to get better until you agree on how you’re doing. Specific remedies are left to state organizations and policymakers, who can dig into each state’s individual report to see where it leads or lags its neighbors.

The review, published in Advances in Nutrition, splits the evidence into two unequal halves. On one side sit 48 controlled trials, mostly testing single nutrients in adolescents. On the other sit 25 prospective studies that tracked children’s diets from infancy and waited, sometimes nineteen years, to see how their minds turned out.

It is the waiting studies that carry the weight. Across cohorts in Australia, the UK, the Netherlands and China, a consistent signal surfaced: the quality of a baby’s diet, especially in the first twelve months, predicted intelligence scores years later, well into the school years and adolescence. Higher intake of fruit, vegetables, dairy and whole grains tracked with stronger verbal and full-scale IQ. Diets heavy on processed and sugary foods tracked the other way. And the association was strongest for diet at age one, fading for diet at two and three.

A window that may have already closed

Why the first year? The brain is doing something extraordinary then. Total brain volume roughly doubles in those twelve months, an increase of around 101 per cent, against a mere 15 per cent in the second year. Glucose metabolism in the frontal cortex surges, myelination sweeps across the brain. A period of construction on that scale is also, inevitably, a period of vulnerability. Skimp on the raw materials and the building may differ.

The harder finding concerns iron. Three studies followed children who had iron-deficiency anaemia as infants and were then treated, their blood counts restored to normal. By the time these children reached 10, or 14, or 19, their iron status looked fine. Their brains did not behave as though it were. They showed poorer inhibitory control, weaker executive function, smaller amplitudes in a brain-wave marker called the P300, as though an early shortfall had left a watermark that later correction could not bleach out.

“The foundations of cognitive health appear to be laid very early in life, and the effects can still be seen in adolescence,” says Professor Hayley Young, who led the review.

So is the case closed, the damage all done before a child can walk? Not quite, and this is where the trials in teenagers come back in. Adolescence is itself a second great rebuild: synaptic pruning, more myelination, the slow maturation of the prefrontal cortex that governs planning and self-control. All of it driven, in part, by the hormonal storm of puberty. If the brain is plastic again, it should be sensitive again, and the question becomes whether you can still nudge it.

The trouble with feeding the well-fed

Here the trials get messy, and the mess is instructive. Iron helped, but mainly in adolescents who were already deficient or anaemic, lifting verbal memory and nonverbal IQ. Iodine improved nonverbal reasoning in deficient teenagers, a “catch-up” effect, but only if their iodine levels were genuinely restored; one trial that failed to top them up properly found nothing. Omega-3 fatty acids were the great disappointment, with no consistent benefit, though the authors note that some trials barely shifted the participants’ omega-3 levels at all. The Food2Learn study is the cautionary tale here: a fifth of participants dropped out, a third stopped taking the capsules, and the omega-3 index crept up by just over one percentage point across an entire year.

The pattern, once you see it, is hard to unsee. The nutrient tends to help the child who lacked it, and does little for the child who did not. Feed iron to a replete teenager and you are pushing on a door that is already open. This is awkward for the dream of a brain-boosting supplement aisle, but it is exactly what you would expect if these nutrients work by fixing deficits rather than supercharging the typical brain.

There is a further wrinkle, and the authors are candid about it. Only two of the infancy studies controlled for the mother’s IQ, and verbal intelligence, the very thing most strongly linked to early diet, is also the thing most shaped by a child’s home and schooling. A clever, health-conscious parent feeds a child well and also fills the house with books and conversation. Untangling the diet from everything that travels with it is genuinely hard, and the review does not pretend otherwise. (The work was funded by an industry-backed nutrition body, a detail worth keeping in view, though the conclusions are notably restrained rather than promotional.)

What the Swansea team offers instead of a headline is a method. Their reading is that the contradictions are not noise to be averaged away but signal in disguise. As the paper puts it, the impact of diet depends on “who is studied, when in development exposure occurs, what is delivered … which domains are assessed, and the context in which interventions are implemented.” Study iron in the iron-replete and you find nothing; study it in the anaemic and you find something. Both results are true.

To make future studies less of a muddle, the authors lay out seven principles: track diet across the whole life course rather than one snapshot, study whole diets instead of lone nutrients, measure biomarkers to confirm the stuff actually got in, account for puberty and sex, standardise the cognitive tests, attend to poverty and food access, and control properly for confounders. It reads less like a conclusion than a confession that the field has been doing it the hard way.

The open question, Young says, is “whether adolescence itself is a second window of opportunity.” Mild iodine deficiency is quietly re-emerging in wealthy countries, the UK and US among them, as eating habits shift. If the teenage brain really is listening, even a little, then what a 14-year-old eats might still matter for the adult mind they are building, long after the first thousand days are spent. Nobody yet knows. That, rather more than any single nutrient, is what the next round of studies will have to settle.

DOI / Source: https://doi.org/10.1016/j.advnut.2026.100648

Frequently Asked Questions

Does eating well as a teenager actually make you smarter?

The honest answer is: it depends who you are. Across dozens of trials, dietary improvements mostly helped adolescents who were already short of a key nutrient like iron or iodine, and did little for those who were already well-nourished. There is no good evidence that piling supplements onto a healthy teenager boosts cognition, but correcting a genuine deficiency can produce real gains.

Why does the first year of life keep coming up?

Because the brain grows faster then than at any later point, with total brain volume roughly doubling in twelve months. That breakneck construction makes the first year both a window of opportunity and a period of vulnerability, which is why diet at age one predicted later intelligence more strongly than diet at two or three in the studies reviewed.

If a child’s iron deficiency is treated in infancy, is the problem solved?

Not necessarily. Several long-term studies found that children treated for iron-deficiency anaemia as babies still showed differences in executive function and brain activity at ages 10, 14 and even 19, despite having normal iron levels by then. It suggests early shortfalls can leave lasting marks that later treatment does not fully erase.

Is it true that omega-3 supplements don’t help teenagers?

The review found no consistent benefit, but that may be partly a story of failed delivery rather than a failed idea. In several trials, participants stopped taking the capsules or their omega-3 levels barely rose, and benefits tended to appear only when those levels climbed past a certain threshold. The nutrient’s value for the typical adolescent brain remains genuinely unsettled.

The work, led by Yuanyuan Diao and colleagues at the Beiersdorf Innovation Center Shanghai, set out to do something the cosmetics industry has wanted for years: turn the slippery idea of “looking your age” into a number you can actually measure. And measure reliably, with ordinary people doing the looking rather than a trained dermatologist or a machine.

Perceived age, as the field calls it, is an old notion. Francis Galton was poking at it back in the 1880s, noticing that some people simply read as older or younger than their birth certificates would suggest. For a long time it stayed a curiosity, a thing everyone noticed and nobody quantified. Only with standardised photography, and then digital imaging, did it harden into something a study could pin down. The Shanghai team wanted to take the next step and validate a version of it that leans on lay judgement, the kind of judgement that actually operates in the world when someone glances at your face across a table.

How You Build a Number Out of a Glance

The setup was meticulous, almost clinical. Chinese women aged 15 to 65, evenly spread across ten five-year bands, were photographed under controlled light with their hair tucked away and every accessory removed. A dermatology technician scored their faces against the Asian Skin Aging Atlas, grading wrinkles and folds and sagging on fixed scales. Then, on a separate visit, the women sat a metre from a 27-inch screen and guessed the ages of strangers from the photo set, working only from faces in age bands adjacent to their own.

Why neighbours rather than a single panel of experts? Because people read faces of their own generation more accurately, and stacking sixty-odd independent guesses per face and taking the median washes out the wild outliers. The average woman, it turned out, was judged 1.6 years older than her actual age. A small number on its own. But the average hides the interesting bit.

The young got the worst of it. Teenagers between 15 and 20 were pegged, on average, more than five years older than they were. The oldest group, 61 to 65, went the other way and were read as 2.4 years younger. The researchers describe perceived age as an integrated, holistic measure rather than a tally of separate flaws, which is part of why it tracks biological aging so well.

The Folds That Give Us Away

So what is the eye actually catching? Across every group, two features dominated: the nasolabial fold, the crease running from nose to mouth corner, and the marionette fold below it. These topped the list whether a face was being aged up or down. After them came the softer, less obviously “wrinkle” cues, the tightness of the facial contour, the evenness of skin tone, and radiance. The things, in other words, that no single line on a face can account for.

The statistics carry a sting for the young. In the teenage group, overall facial sagging and under-eye lines more than doubled the odds of being judged older, presumably because any blemish on an otherwise smooth, even, firm young face stands out as something gone wrong, an anomaly the eye reads as premature aging. For older faces the logic flips: assessors lean on whether the prominent markers are present or absent, so a sixty-something who has kept her skin even and bright can read as unusually well preserved. Dullness and yellowness, meanwhile, pushed older faces older still, in the oldest band raising the odds of overestimation by more than two-thirds.

Crow’s feet, oddly, barely registered, ranking 17th of 25 features here, even though a separate multi-ethnic study had flagged them as a major tell in Chinese women. Faces, clearly, do not hand over their secrets in quite the same way to every set of eyes.

There are caveats worth keeping in view, and the authors are candid about them. The whole cohort came from a single city, Guangzhou, with one climate and one broad skin phenotype, so the numbers may not travel to other regions, never mind other ethnicities. There is also the awkward fact that the assessors knew they were in an aging study, which could have nudged them to scrutinise faces more harshly than they would in a café. And the study was paid for by Beiersdorf, the company behind Nivea and Eucerin, whose interest in a quick, cheap, consumer-friendly aging yardstick is not exactly disinterested.

Still, the practical pull is obvious. A trained grader needs roughly 30 seconds per facial parameter; a perceived-age guess takes about three. Push the photos online and you can gather a crowd of assessors in an afternoon rather than booking a clinic for three months. For a company wanting to know whether a cream actually makes someone look younger, rather than whether it nudged a wrinkle score by a decimal, that is a tempting shortcut.

What lingers, though, is the smaller, stranger finding buried in the data. We are hardest on the young, quickest to mistake a single early line for the onset of age, and oddly generous to the old. The face we present to the world is read not against some absolute scale but against what a stranger expects a face like ours to look like. Whether the next wave of skincare can actually shift that verdict, or merely measure it more cheaply, is the question the industry has yet to answer.

DOI: 10.1016/j.jdsct.2026.100148

Frequently Asked Questions

Why are young people judged as older than they are?

On a smooth, even, youthful face, any small sign of aging stands out sharply against what observers expect, so the eye over-weights it. In this study, facial sagging and under-eye lines more than doubled the odds that a teenager would be guessed older than her real age. The same features on a sixty-year-old barely move the needle, because they are already expected.

What facial features most strongly drive how old we look?

Two folds dominate across all ages: the nasolabial fold from nose to mouth, and the marionette fold beneath it. After those, observers rely on the tightness of the facial contour, the evenness of skin tone, and skin radiance, all features that no single wrinkle can capture. Crow’s feet, interestingly, ranked low here despite being prominent in some other studies.

Is perceived age actually a reliable measurement?

In this cohort it tracked closely with expert-graded aging features, and pooling more than sixty independent guesses per face and taking the median smooths out individual error. That said, the assessors all came from one city and knew they were in an aging study, so some bias is hard to rule out entirely. It is best treated as a practical complement to clinical grading rather than a replacement.

Could this change how anti-aging products are tested?

That is the explicit hope of the team behind it. A perceived-age judgement takes about three seconds against roughly thirty for a single clinical parameter, and photos can be assessed by large online panels rather than in a clinic. It also measures what consumers actually care about, looking younger overall, rather than an abstract wrinkle score.