Mind & body, in brief

In obstructive sleep apnoea the upper airway narrows or shuts again and again during sleep. Each blockage usually lasts about 30 seconds and ends with a drop in oxygen, a brief waking, or both. [1]

The collapse comes from the throat muscles losing their tone in sleep, and from the shape of the airway itself. [2] About 70% of people with apnoea also wake too easily, have weak muscle responses, or have breathing control that overreacts. [1]

Repeated blocked breaths set off the body's alarm system, which raises blood pressure and makes the heart work harder. [2] About 1 billion adults worldwide are estimated to have apnoea. [1] Across seven studies of 594 men, more blocked breaths an hour went with worse erectile function. [11]

Extra body weight accounts for an estimated 40% to 60% of cases, partly through fat inside the throat and neck. [7] In 538 people who declined treatment, apnoea usually got worse over about four years, and faster in those who were heavier or gained weight. [10]

The standard treatment is CPAP, a machine whose pump pushes air through tubing and a mask to hold the throat open. [2] Across 136 trials with 8,827 patients, people used it for 4.5 hours a night on average. [4]

Only 7.8% of those trials reached six hours of use a night. [4] Up to half of patients stop using CPAP within the first year. [3] The reasons people give include sore marks from the mask, air leaks that dry the eyes, and feeling closed in. [2]

A mouth guard that pushes the lower jaw forward stops fewer blocked breaths than CPAP. It can do about as much for sleepiness and blood pressure, probably because people wear it more. [3]

The largest trial of CPAP against heart attacks and strokes, called SAVE, found no benefit on average. [9] Its 2,687 patients were re-analysed in 2026 with a computer model, which suggested the mask may help some people and harm others. [9]

A new pill, AD109, pairs two drugs. Atomoxetine helps switch on the nerves to the throat muscles in sleep, and aroxybutynin blocks a signal that slackens those muscles in dream sleep. [3][2]

In a 26-week trial across 69 centres, 646 people who could not or would not use CPAP took AD109 or a dummy pill. Blocked breaths fell by 44.1% on the pill and by 17.6% on the dummy. [5]

Tiredness did not improve more on AD109 than on the dummy pill. [5] Across 13 trials of atomoxetine combinations, dry mouth hit 29.2% of people, against 9.5% on dummy pills. [6]

In the US, the weekly weight-loss injection tirzepatide is the only drug approved to treat the blocked airway itself, and only in adults with obesity. [3][7] In its trials, the fall in the top blood pressure number was tied to weight loss, and not to fewer blocked breaths on their own. [8]

Why it matters. AD109 is still an experimental drug that needs a regulator's approval, so the next test is whether people keep taking a nightly pill that one in five quit over side effects. [3][5]

The US Food and Drug Administration approved daraxonrasib, sold as Rasonque by Revolution Medicines, on 26 August for advanced pancreatic cancer. [12] It shuts off a faulty gene called KRAS, which drives growth in more than 90% of pancreatic tumours. [13] In a 500-patient trial, people taking the daily pill lived a median 13.2 months, against 6.7 months on chemotherapy. [13] Severe side effects hit 44% on the pill and 57.5% on chemotherapy, and the agency approved it six months ahead of its deadline. [13]

Why it matters. Pancreatic cancer has the highest death rate of the major cancers, and more than half of patients die within three months of diagnosis. [13] It is the first approved drug to attack a genetic cause of the disease. [12]

On 11 September the US Food and Drug Administration approved Isembyld, made by Scholar Rock, for spinal muscular atrophy, a rare disease of the nerves that control movement. [14] Existing treatments target SMN2, a gene those nerves depend on. [14] Isembyld is approved for adults and children aged two and over who already take one of them. [14] In the late-stage trial, young patients' movement skills improved over a year, while those on a dummy drug declined. [14]

Why it matters. It is the first approved therapy aimed at the loss of muscle in the disease. [14] The company says decades of attempts to block myostatin, a protein that limits muscle growth, had failed before this one. [14]

In type 1 diabetes the body's immune system destroys the cells that make insulin. [16] A 2026 review reports that zimislecel, made by Vertex from stem cells, let 10 of 12 people in an early trial stop injecting insulin. [15] Those patients had suffered dangerous drops in blood sugar. [15] In the larger FORWARD trial, every participant spent more than 70% of the time with blood sugar in the target range after one infusion. [16] They also had to take drugs that suppress the immune system. [16]

Why it matters. Donor pancreases are scarce and vary in quality, and cells grown in a lab could be made to a standard. [15] The need for immune-suppressing drugs is what still limits who can have them. [15]

Islets are the clusters of cells in the pancreas that make insulin, and they can be taken from a donor and infused into a patient's liver. [17] A 2026 review reports that in one study of 36 recipients across nine centres, 58% became free of insulin injections. [17] By two years, 76% of them were back on insulin. [17] The review says the cells die off from lack of oxygen, and patients need lifelong immune-suppressing drugs. [17]

Why it matters. Researchers are developing coatings that shield transplanted cells from the immune system, to get around both problems. [17]

Researchers at the Murdoch Children's Research Institute in Australia analysed 106,581 infection samples from children in 82 countries. [18] Resistance to antibiotics rose in every region between 2004 and 2022. [18] Babies and children in intensive care, and children in poorer countries, were hit hardest. [18] The team projects resistance to last-resort antibiotics will reach 82% for the bacterium Acinetobacter baumannii and 35% for Klebsiella by 2035. [18]

Why it matters. Children get many bacterial infections and have fewer antibiotic options than adults, with few medicines made in child-friendly forms. [18] The researchers found resistance even in simple urine infections, not only in severe illness. [18]

The World Health Organization gathered records of more than 23 million bacterial infections from 104 countries. [19] It found one lab-confirmed infection in six was resistant to antibiotics in 2023. [19] More than 40% of the antibiotics tracked lost strength against common blood, gut, urine and sexually transmitted infections between 2018 and 2023. [19] The rate was one in three in south-east Asia and the eastern Mediterranean. [19]

Why it matters. In 2021, drug resistance played a part in 4.71 million deaths and directly caused 1.14 million. [19] The head of its resistance department said treatment options are running out, especially where infection control is weak and lab tests are hard to get. [19]

A study in the journal Nature Medicine scored how 193 countries governed antibiotic resistance from 2017 to 2022, using 269 policy documents. [20] The average score rose from 30.7 to 44.5 out of 100. [20] Carrying the plans out, and checking on them, lagged behind writing them, especially for animals and the environment. [20] A measurable change in resistance appeared only about five years after a country adopted its plan. [20]

Why it matters. The global action plan on resistance is being updated in 2026. [20] The study found that working across health, farming and animal sectors, and tracking antibiotic use, were most linked to better results. [20]

Cochrane, an international group that reviews medical evidence, published a review in May of six trials with 800,000 men in Europe and North America. [21] It found PSA blood testing likely prevents about two prostate cancer deaths for every 1,000 men screened. [21] Its 2013 review had found no such benefit, and PSA testing fell out of favour after it. [21] The new review had up to 23 years of follow-up. [21]

Why it matters. The benefit is small, and the test also finds slow cancers that would never have caused harm, whose treatment can cause infections and impotence. [21] The authors said the finding is not a blanket case for screening every man. [21]

US researchers followed 1,039,659 veterans who got a flu vaccine in late 2024, of whom 349,085 also got that season's COVID vaccine. [23] Over eight months, the COVID vaccine was linked to a 37.7% lower risk of heart attacks, strokes and heart deaths tied to COVID. [23] The benefit was clear only in people over 75. [23] The vaccinated also had about 24% fewer heart events of any kind. [22]

Why it matters. The lead researcher thinks many of those extra heart events came from COVID infections nobody tested for. [22] The study used health records rather than assigning vaccines by chance, so it shows a link. [23]

France's national health data covered 22.8 million adults aged 18 to 59 who got an mRNA COVID vaccine in 2021, and 5.9 million who did not. [24] Over about 45 months, the vaccinated had a 25% lower risk of dying from any cause. [24] Their risk of death from severe COVID was 74% lower. [24] The study found no increase in deaths from any cause in the vaccinated over four years. [24]

Why it matters. No earlier study had compared long-term deaths by vaccination in young adults, who are less likely to get severely ill. [24] It is a cohort study weighted for 41 other illnesses, not a trial that assigned vaccines by chance. [24]

Dutch researchers looked at 75,421 people who received COVID vaccines and died between January and November 2021. [25] Their average age was 79.8. [25] Deaths were 44% less frequent in the three weeks after a dose than in later periods. [25] Deaths were 16 times more frequent after a registered positive COVID test. [25]

Why it matters. Each person who died is compared with themselves at other times, which cancels out lasting differences such as frailty and other illnesses. [25] In late 2021 Dutch deaths ran above what COVID deaths alone explained, which raised the question of other causes. [25]

A French trial randomly split 1,160 healthy people, average age 80, who took statins to prevent heart disease. [26] One group stopped their statins and the other kept taking them, for three years. [26] In the group that stopped, bad cholesterol rose 50% within three months. [26] After three years 7.2% of the stoppers had died, against 7.9% of those who continued, and heart attacks and strokes did not differ. [26]

Why it matters. Earlier studies in Denmark and Italy found about 30% more heart trouble after stopping, but they did not record why people stopped. [26] A US cardiologist said the result will not change care for most people over 75. [26]

Psilocybin is the drug in magic mushrooms. The EPISODE trial at two German centres enrolled 144 people whose depression had resisted other treatment. [27] With therapy sessions, they got 25 mg of psilocybin, 5 mg, or nicotinamide, a vitamin used as the comparison. [27] After six weeks, 17.0% responded on 25 mg, 12.5% on 5 mg and 10.6% on the vitamin. [27] The difference was not statistically significant. [27] Thoughts of suicide on dosing days were reported by 4% on 25 mg, against 1% to 2% otherwise. [27]

Why it matters. Because its main comparison failed, no further formal tests were run. [27] Two serious reactions followed the 25 mg dose, including one case of lasting visual disturbance. [27]

A Stockholm trial gave 35 people with recurring depression one 25 mg dose of psilocybin or niacin, a vitamin used as the comparison, plus five support sessions. [28] Depression scores were better on psilocybin at day 8, day 15 and day 42. [28] By day 365 the difference was no longer significant. [28] Two people on psilocybin had severe, lasting anxiety that needed medical care. [28]

Why it matters. Psilocybin has been put forward as fast-acting with lasting effects, and here people's own ratings were better from day 2. [28] Those self-ratings stayed better until about day 102. [28]

A trial at one NHS site in England gave 60 people with hard-to-treat depression 25 mg of psilocybin or a dummy, with preparation and support. [29] It was designed to test whether such a trial can be run, and 59 of 60 completed every follow-up. [29] At week 3 depression scores were 10.41 points better on psilocybin, and the gap held at week 6. [29] Non-serious side effects numbered 164 on psilocybin and 123 on the dummy. [29]

Why it matters. It tested whether psilocybin therapy can run inside a public health service. [29] The authors say the results support a larger confirming trial. [29]

In February, Compass Pathways released results from two late-stage trials of its psilocybin drug, COMP360, for severe depression. [30] Patients on the drug improved more on a depression scale than the comparison group. [30] It could become the first psilocybin medicine on the market, and the second psychedelic after Spravato, a ketamine-based spray. [30] A US psychiatrist not involved said the data probably meet the bar for approval, without being miraculous. [30]

Why it matters. The company released the headline in a press statement, and fuller data are needed to judge how much patients gain. [30] Psilocybin could be approved in the US as early as next year. [31]

Oregon lets most adults over 21 take psilocybin at licensed centres with a supervisor and no prescription. [31] A study published on 19 August followed almost 350 of the more than 20,000 people who have gone through the programme. [31] More than half called it one of the most meaningful experiences of their lives, and four had serious reactions. [31] Three reported new thoughts of dying or suicide a week later, and that number doubled by three months. [31]

Why it matters. It is the most rigorous look yet at psilocybin outside a medical trial. [31] About a third of participants had household incomes above 200,000 dollars, and the researchers say the low number of bad reactions may partly reflect who chose to join. [31]

Obsessive-compulsive disorder, OCD, involves intrusive, unwanted thoughts and compulsive behaviour. [33] A small trial gave 15 people with OCD weekly psilocybin sessions, and some got a sedative as a comparison. [32] After eight weeks and at least four high doses, 73.3% had responded and 40% were in remission. [32] The effect shrank but stayed substantial at six months. [32] A review found 40% to 60% of OCD patients get partial or no relief from existing treatment. [33]

Why it matters. The same review found stronger signs of benefit for psilocybin than for cannabis, though it had to include unpublished results. [33] The trial's main aim was safety, and it had only 15 people. [32]

The American Psychological Association reports that people with OCD have symptoms for nearly 13 years, on average, before a correct diagnosis. [34] OCD affects nearly 10 million people in the US, and more than 80% of cases are estimated to be undiagnosed. [34] As few as 2% of those diagnosed may be getting the recommended treatment. [34] That treatment, exposure and response prevention, teaches people to live with the thoughts without acting on them. [34]

Why it matters. People with OCD face a suicide risk five times that of the general population. [34] Frightening intrusive thoughts are sometimes mistaken for a sign of danger to self or others, leading to hospital stays. [34]

Ondansetron is a drug that blocks one type of receptor for serotonin, a chemical messenger. [35] A review pooled five trials with 289 adults whose OCD persisted despite standard treatment. [35] Adding ondansetron improved symptom scores more than a dummy pill at week 4, week 8 and week 12. [35] At week 8 the gain was 5.53 points on the standard OCD scale. [35] The trials were small, and results varied between them. [35]

Why it matters. A large share of adults with OCD keep their symptoms despite the best first-line treatment with therapy or antidepressants. [35]

Cognitive behavioural therapy teaches people to change how they respond to thoughts and fears. A review of 61 trials with 3,710 people with OCD compared ways of delivering it. [36] Face-to-face, remote, guided self-help, intensive and family-involved therapy all worked about equally well. [36] Group therapy did less than individual therapy, and self-help done alone did less than most formats. [36] More people dropped out of self-help done alone. [36]

Why it matters. How therapy was delivered mattered less than whether anyone guided it: nearly every format beat self-help done alone. [36]

Acceptance and commitment therapy teaches people to make room for difficult thoughts rather than fight them. A review pooled nine trials of it for OCD, from the US, Denmark, Iran and South Korea. [37] It worked moderately better than no treatment or drug-only comparisons. [37] It did no better than other talking therapies. [37] Once the comparisons were matched, it worked about equally in Eastern and Western studies. [37]

Why it matters. Five of the nine trials came from Iran, so the evidence leans heavily on one country. [37]

An umbrella review brought together 28 earlier reviews covering 24,762 young people under 18 with OCD. [38] Cognitive behavioural therapy, especially exposure and response prevention, had the largest and longest-lasting effects. [38] Antidepressants of the SSRI type worked moderately, and combining them with therapy may help in severe cases. [38] Therapy delivered over the internet was also reviewed. [38]

Why it matters. The authors say treatment needs to become easier to reach and better suited to children's ages. [38]

Studies reported micro- and nanoplastics in brains, testes, placentas and arteries. [39] In January the Guardian identified seven such studies challenged by other researchers in the same journals. [39] One analysis listed 18 studies that did not account for human tissue giving signals that look like common plastics. [39] A researcher at a German environmental research centre said fat gives false readings for polyethylene, and the brain is about 60% fat. [39]

Why it matters. The brain study's senior author said the criticism was mostly not based on data. [39] The BMJ, a medical journal, reported in March that new research suggests plastic levels may have been overestimated. [40]

A review of 26 studies with 564 human samples found counts from under one to several thousand particles per gram of tissue. [41] It says cleaner, plastic-free lab methods are lowering the overestimates. [41] A second review of 190 measurements found the opposite for solid organs. [42] It says automated machines found nearly 70 times more particles in tissues like placenta and lung than older hand methods. [42]

Why it matters. Both reviews blame the measuring methods, and they point in opposite directions. [41][42] The first review calls for standard lab methods, because studies done in different ways cannot be compared. [41]

A review in the journal Communications Medicine looked at micro- and nanoplastics and cancers of the prostate, bladder and kidney. [43] Some studies have measured particles inside prostate and bladder tumours, with higher amounts linked to eating take-out food often. [43] Suggested ways they could cause harm include inflammation and chemicals added to plastics. [43] The review says the evidence is not yet enough to link them directly to these cancers. [43]

Why it matters. The same particles are being tested as carriers for cancer drugs, so the review covers both a possible harm and a possible use. [43]

Researchers at Ohio State University compared 256 adults under 50 with lung cancer against 2,921 people without cancer. [44] Those who only smoked had about five times the odds of lung cancer. [44] Those who both smoked and vaped had about 13.8 times the odds. [44] Most of the young people's tumours, 72%, were adenocarcinomas, a type that starts in the lung's gland cells. [44]

Why it matters. It is a case-control study, which compares the past habits of people with and without the disease, so it shows a link rather than proof. [44] Lung cancer is the leading cause of cancer death worldwide, including in young adults. [44]