For how we live.
Mental health, psychedelics, loneliness, and neuroscience.
A single-centre randomised trial found the Bergen 4-day treatment cut obsessive-compulsive symptoms far more than an unguided internet CBT programme, with 61% responding at three months versus 25%.
A prespecified analysis of 227,502 veterans who contacted the Veterans Crisis Line found suicide mortality of 0.29% among caring-letter recipients versus 0.27% in historical controls — no measurable difference.
Ketamine cuts suicidal thinking fast but briefly. A 50-person trial gave low-dose buprenorphine afterward and saw the effect hold — an early, single-centre result in a high-risk group.
A Global Burden of Disease 2023 analysis of 31 countries finds the burden climbing, peaking in early adolescence and heavier in women, with population growth the single biggest driver.
There is no approved drug for methamphetamine use disorder. In 339 patients treated in ordinary clinics, a generic antidepressant cut use by about two days a month more than placebo, with more dropouts from side-effects.
In LatAm-FINGERS, 1,065 at-risk older adults across 11 countries did better on cognitive tests with structured coaching than with flexible health advice over two years. The gap was small.
In the industry-funded trial, young responders who kept taking the dopamine D1 blocker relapsed less than those tapered to placebo. The effect was clear in children but not in the small adult group.
In two randomised trials, a grief-specific psychotherapy roughly doubled response rates against strong comparators. Citalopram alone did not beat placebo — but it did help the depression that often travels with grief.
Anxiety disorders in young people are treatable. In the largest trial, combining CBT with an SSRI worked best, and both beat placebo. Talking therapy alone also clearly outperforms no treatment.
The largest synthesis of experimental sleep research finds that shorter, broken or lost sleep flattens positive emotion and ramps up anxiety — and trials that fix sleep improve mental health in turn.
Across 154 studies, activities that lower arousal — slow breathing, meditation — reduced anger and aggression. Venting-style activities that raise it, like hitting a bag or going for a run, did not.
At high doses caffeine reliably provokes anxiety and can trigger panic attacks in vulnerable people. The effect is dose-dependent, habitual drinkers differ, and disrupted sleep is part of the story.
Slow breathing does shift the body toward calm, but the popular 4-7-8 count has almost no direct trial evidence, and the breathwork that has been tested in a large study is a different pattern.
Repeating 'I'm a lovable person' left people with low self-esteem feeling worse in controlled experiments. A different technique — writing about your core values — has stronger evidence behind it.
A 40-person Harvard trial aimed accelerated magnetic stimulation using each patient's own connectivity scan. Depression scores fell further than with standard scalp measurement, though the trial was small.
Retrospective studies keep finding bilinguals develop dementia symptoms three to four years later than monolinguals. Prospective studies that follow healthy people find no drop in who gets it at all.
Bigger weekly doses of nature track with better wellbeing, structured outdoor programmes ease depression and anxiety in trials, and a single walk can quiet the overthinking that feeds low mood.
Rumination — replaying the same worries on a loop — is a genuine risk factor for depression, and the best trials show it can be trained down, though no single technique clearly beats the rest.
A network meta-analysis of 54 trials put behavioural and cognitive therapies ahead of the drugs on OCD symptoms. But most therapy trials also allowed antidepressants, hinting the two work best together.
US POINTER randomised 2,111 at-risk older adults to two lifestyle programmes. The structured version beat the self-guided one on cognition — but by a modest margin, and there was no do-nothing group to compare against.
A 43-year meta-analysis found loneliness in young adults rising by about half a standard deviation — while its own authors call the word 'epidemic' exaggerated. The pandemic added a small further bump.
The best single randomised trial found morning light beat a sham device over eight weeks, but the effect is most reliable when light is added to an antidepressant rather than used alone.
A Cochrane network meta-analysis compared eight talking therapies for panic disorder. Cognitive behavioural therapy came out ahead most often — but the effects were small and the evidence graded low throughout.
A gratitude habit gives most people a genuine but small lift in wellbeing — a 2025 meta-analysis of nearly 25,000 people put the average effect at a modest 0.19 — and its unique benefits are probably overstated.
The idea that sleep flushes toxins from the brain rests largely on mouse experiments. A 2024 study using the same animal measured the opposite direction, and none of it has been shown in humans.
Pooled trials put the herb roughly on a par with standard antidepressants for mild-to-moderate depression, with fewer dropouts for side effects. The catches are dangerous drug interactions and wildly variable products.
The largest network meta-analysis of social anxiety disorder ranked 41 options against each other. Individual CBT came out on top, and it was one of only two treatments to clearly beat a placebo.
Everyday lapses are a poor guide to trouble. The line that matters is whether memory loss is measurable beyond your age and whether it is starting to affect daily function.
Two of the largest datasets — 6.5 million UK bank customers and a Swedish national registry — tie heavier gambling to financial and health decline, and gambling disorder to a fifteenfold suicide death rate.
Compulsively scrolling bad news tracks with worse mental and physical health — but the strongest studies show association, not proof, and the scariest prevalence figure shrank by half on replication.
In a 619-patient NEJM trial of treatment-resistant depression after 60, augmenting an antidepressant with aripiprazole improved well-being more than switching to bupropion. The margin was modest.
Older meta-analyses linked lithium to fewer suicides. A randomised trial in 519 US veterans, stopped early for futility, found no such benefit when it was added to usual care.
The largest network meta-analysis of brain stimulation found several forms of TMS roughly triple the odds of responding, but they sit below ECT and rest on mostly small, uncertain trials.
In the trial that pitted them against each other, aerobic exercise matched sertraline on remission. But both sat only a little above a strikingly strong placebo, which is the part the headlines skip.
Cobenfy, approved in the US in September 2024, cut symptom scores more than placebo in two five-week trials — without the weight gain and movement effects of older drugs, but with cholinergic side effects.
A US trial found ketamine at least as good as electroconvulsive therapy; a Swedish trial found ECT clearly better. The gap lies in whom they treated and how they scored success.
In the largest head-to-head trial, a simpler therapy delivered by junior staff matched full CBT — mean difference 0.1 PHQ-9 points. A Cochrane review finds no difference between the two.
A matched cohort of 173,910 people with dementia in England tied the drugs to raised risks of pneumonia, stroke, clots, heart attack, fracture and kidney injury — highest just after starting.
In the largest network meta-analysis, EMDR and trauma-focused CBT were the most effective treatments for adults with PTSD. The evidence is graded moderate-to-low, and the eye movements may not be why it works.
Dopamine is a learning signal that tracks surprise, not a pleasure tank you drain and refill. Stripped of the neuroscience label, the practice is ordinary stimulus control — and that narrower version has some support.
Restrictive English schools cut phone use during the school day but not overall — and teenagers' wellbeing was no better than at schools that allow phones. What tracks with worse mental health is heavy, problematic use.
The largest meta-analysis puts the incidence attributable to the drug at about 15% — one in six or seven. Critics who reanalysed part of it put it above 50%. The gap is about what counts as evidence.
Registry and responder cohorts document persistent PTSD and depression, a symptom peak more than a decade after the attacks, and a burden that compounds with chronic physical illness.
A meta-analysis of consumer neurofeedback found no benefit for cognition, mindfulness or physical health and only a small effect on distress — even though the headbands do record real brain activity.
A Danish study of 71,000 cases found the risk factors tied to a diagnosis weakened over a decade, pointing to changing diagnostic practice rather than falling thresholds of impairment.
A British Columbia study emulated a trial across 41,000 patients. Those started on take-home methadone within weeks to months of induction stayed in treatment longer and died less than those on daily witnessed dosing.
An arthritis antibody and a diabetes drug were given to depressed patients in 2026 trials. Both were designed to probe a mechanism rather than prove a treatment, and reading them as results would be a mistake.
A subculture built on ranking faces sits directly on top of a recognised, treatable mental illness. The distress, not the jawline, is the part the evidence says to take seriously.
Two 2026 trials point the same direction: the hardware in depression treatment is mature, and the remaining gains are in precision and in what the treatment costs the patient's memory.
Extreme-leanness culture rebrands restriction as discipline. The evidence is a mental-health story: eating disorders carry among the highest mortality in psychiatry, and under-fuelling harms the body first.
The best evidence frames procrastination as a way of dodging bad feelings about a task, not a scheduling flaw — and following students over a year, more of it tracked with more depression, anxiety and stress.
Among 383 dialysis patients, 83.6% screened for moderate-to-extreme anxiety and 58% for PTSD; a separate study of 250 displaced adolescents found trauma exposure tracking closely with post-traumatic symptoms.
Structured counselling for a spouse delayed a patient's nursing-home admission by an estimated 557 days. Other trials cut carer depression sharply, though the effect on institutionalisation is less consistent.
A Cochrane review of 79 trials and 24,308 patients found the primary-care model helps with depression and anxiety, modestly in the short and medium term and not at all in the very long term.
In 415,000 New Jersey births, infants of mothers who screened positive for depression died before age one at more than three times the rate, an association steady across race, income and preterm birth.
A German trial in 1,793 students found a cognitive-behavioural programme reduced gaming and internet-use disorder symptoms only in high-risk adolescents — not across the whole classroom.
The most popular cognitive-enhancer stack has one modest, well-tested combination at its heart. Beyond caffeine and L-theanine, the evidence in healthy people thins out fast.
In the Ket-BD trial, four IV ketamine infusions over two weeks lowered depression scores more than an active placebo, with no manic switches observed in 63 analysed patients.
A network meta-analysis of 163 trials found EPA, vitamin D and a St John's wort extract beat placebo by margins so large the researchers treat them as a warning about the evidence, not a green light.
Dronabinol beat placebo on a nightmare-severity scale in 171 adults, but the effect was moderate and the only serious adverse events occurred in the drug group. Long-term safety is untested.
Loneliness and isolation raise mortality risk about as much as well-known risk factors. But trials of interventions to relieve them in older people are mostly low-quality, and the best do more than arrange company.
Two popular stories — alcohol as a cause of low mood, and drinking as self-medication for it — both get less support from recent genetic and within-person data than the confident telling suggests.
A survey of seven countries found the youth mental health app market dominated by paid journaling and meditation apps. The AI-powered ones show early promise on symptoms — and none had checked for harm.
In a 40-person crossover trial, 10 mg and 20 mg of edible THC degraded lane-keeping while whole-blood THC peaked at 1.6 and 3.4 ng/mL — near or under the thresholds police enforce.
Two large re-analyses of the same public datasets found the association is real, negative and very small. A third review found the association is not even consistent between the Global North and South.
A landmark cohort found strained spousal caregivers had a 63% higher death rate over four years. Decades of trials show interventions help — more for a carer's knowledge and mood than for the underlying burden.
In 4,868 clozapine users followed to 2023, excess neutropenia risk concentrated in the first 18 weeks and converged with other antipsychotics by around two years.
The ODDESSI trial ran in 30 clusters across five NHS trusts in England. Time to relapse was unchanged; inpatient admission, re-referral and service-user ratings all moved in Open Dialogue's favour.
A 1.5-million-person records study reports lower mortality with GLP-1 drugs than with SGLT2 inhibitors, and the largest absolute gains in people with serious mental illness. The design cannot prove it.
In 276 adults tested at 60, the differences in GrimAge and DunedinPACE were small, their intervals crossed zero, and most of what was there disappeared once smoking was accounted for.
The MTA trial randomised 579 children and found medication management clearly superior short-term. Its own long-term follow-up found the treatment groups no longer differed by 6 to 8 years.
A ten-person phase 1 study measured whether oral rapamycin reaches the central nervous system in early Alzheimer's. It did not — and several disease biomarkers rose over eight weeks.
WHO's ICD-11 lists burn-out, yet places it among reasons people contact health services, not among illnesses. It is defined strictly as a workplace phenomenon, which shapes who should be assessing it and how.
The Global Burden of Disease estimates for anxiety disorders separate two things that public debate keeps merging: how many people have it, and how likely a given person is to. Diagnosis is a third question again.
In 62 adults scanned under psilocybin during rest, meditation, music, and film, brain networks reorganized differently by setting — and more reorganization tracked with more positive experiences.
Using causal-inference methods on 1,269 people, researchers linked two bacterial genera to depression independent of body weight and medication — the first such evidence in humans, not proof of mechanism.
A 422-family cohort with a pre-ban baseline is the closest thing to a controlled test the world has. Researchers are also warning the law will make studying adolescents harder — including studying this.
A randomised trial tracked infants' cortisol and attachment; a five-year follow-up tracked mental health. Graduated extinction improved sleep with no sign of long-term damage — or lasting benefit.
The TAME trial randomised 276 adults with diagnosed anxiety disorders to mindfulness-based stress reduction or escitalopram. The difference on the primary measure was 0.07 points, inside the noninferiority margin.
Autopsy studies show educated brains carry the same disease but express it as dementia later. That is the concept's strength — and the reason it resists being turned into a to-do list.
The agency compressed review of three psychedelic therapies from a year to weeks. A first approval decision is plausible by early 2027.
Observational studies tie fitness to slower cognitive decline. When researchers randomise older adults to exercise, the cognitive payoff shrinks — in a Cochrane review of healthy people, to nothing.
A network meta-analysis of 331 trials and 34,285 adults found no significant differences between the main psychotherapies, with one exception. In children and adolescents the picture is narrower.
The link between puzzles and a sharper old age is real in the data and largely the wrong way round: the earliest changes of dementia quietly reduce the very activities that later look protective.
The best long-term trial found trained skills can last a decade. Whether they transfer to daily life, or hold off dementia, is where the evidence — and the FTC — parted ways with the marketing.
A network meta-analysis of 101 trials and 11,910 patients found no significant difference between psychotherapy and medication for adult depression — and a consistent advantage for combining them.
Pain medicine added a third mechanistic category in 2017 and has been arguing about it since. There is no test for it, the criteria work by exclusion, and the experts who wrote them say that is a problem.
A standing commission estimates that eliminating 14 risk factors could prevent or delay 45% of dementia across a whole population. What that number means for one person is a different and much weaker claim.
Imaging finds smaller brain volumes from as little as one to two daily drinks. Long-run dementia studies show a U-shape, with abstainers also at higher risk — a pattern that is easy to misread.
A registry study spanning 4.9 million people separates where the most deaths occur from where the inequality is largest. Heart disease and lung disease give different answers.
A review of 110 studies found that 52% of trialled self-guided apps were publicly available — and that evidence quality had no bearing on which ones made it to an app store.
Four separate trial emulations, each with a different comparator, all pointed the same direction. The strongest effect appeared among people being treated for alcohol use disorder.
A review lays out the evidence chain: animal studies showing reduced drug-seeking, one small human trial, and record-based analyses linking GLP-1 prescriptions to 40% fewer opioid overdoses.
A systematic review found only 16 usable studies of how to implement routine symptom monitoring in adult mental health services. A 4,778-patient clinic dataset shows what it looks like when it does.
A meta-analysis of 118 trials found acamprosate and naltrexone reduce drinking, with numbers needed to treat comparable to common heart drugs. They remain widely underused.
A low-probability alert still sends most recipients to a doctor. Researchers are now naming the flip side of continuous self-monitoring — a device-driven health anxiety they call technohypochondria.
Adding the antipsychotic to an antidepressant beat placebo in two phase 3 trials. But the dose that worked in one failed in the other — a wrinkle worth carrying into the prescription.
The KADS extension followed 130 patients who had already been through a ketamine trial. Response fell from 30% at treatment's end to 17% a month later.
The meta-analysis behind ketamine's widening off-label use for suicidal crisis quantifies what 1,166 patients across two decades of trials actually showed — and where the evidence runs out.
A registered report fit identical models to 11 longitudinal studies covering 175,070 people. The subjective feeling carried the signal; the objective count of social contacts largely did not.
A quasi-experimental analysis of 23 hospitals finds about 95 additional self-harm and suicide-related emergency visits a month after the law changed, alongside fewer psychiatric admissions.
A burden-of-disease calculation built on EU noise-mapping data puts numbers on two outcomes that noise research has mostly ignored. The confidence intervals are as informative as the estimates.
A systematic review screened 3,857 records and found three eligible studies, every one produced by a US commercial provider and every one at critical risk of bias.
A month-long trial that blocked mobile internet improved mental health and attention, and a one-week social-media break lifted wellbeing. The wider literature is thinner.
In 273,466 people with substance-related hospital admissions, psychosis was far more common after hallucinogen admissions, until adjustment for preexisting conditions erased the link.
Anxiety, quality of life and PTSD symptoms all improved after a single dose in this 15-person trial. Once researchers accounted for depression improvement, those effects mostly disappeared.
Switzerland allows limited psilocybin use for treatment-resistant depression. A review of 19 patients found meaningful improvement, but remission in barely a fifth to a quarter.
A review pooling 23 studies found risk concentrated among people with bipolar I disorder, family history, or unsupervised use. About 4% later transitioned to bipolar disorder.
A single-author analysis of Global Burden of Disease estimates finds the region's neurodevelopmental picture split in two — and argues the flat autism curve reflects diagnostic capacity, not risk.
By the study's end, 70% had scores low enough to be considered near-resolution. There was no placebo arm — every participant knew what they'd received.
EPISODE randomized 144 people with treatment-resistant depression across four arms. High-dose psilocybin missed its main outcome, even as secondary measures showed symptom benefit.
No medication has proven effective for cocaine use disorder. A single psilocybin dose, paired with therapy, produced more abstinent days and a longer time to relapse than an active placebo.
In 35 patients with recurrent major depression, psilocybin beat an active placebo through day 42. On self-reported measures the effect lingered further, but the trial's primary endpoint showed no difference at 365 days.
TSND-201 improved PTSD severity, disability, and depression scores over placebo in 65 participants — delivered as an oral dose with monitoring but no formal psychotherapy alongside it.
Twenty-four healthy volunteers received all three drugs in a crossover trial. At 30 mg, 2C-B matched MDMA's intensity and produced psilocybin-like states, without psilocybin's anxiety.
The first federal survey to ask about psilocybin specifically finds 2.8% past-year use, concentrated among young, white, college-educated men. People with recent depression had 37% higher odds.
The open-label pilot tested a specific worry in psychedelic research — that borderline personality disorder blunts response to psilocybin for depression. In this small sample, it didn't.
The largest syntheses find apps beat doing nothing on depression and anxiety by about a third of a standard deviation — but the edge shrinks against real treatment, and one umbrella review found no convincing proof.
A survey of 243 people reporting persistent difficulty after a psychedelic trip found that how they responded in the moment, avoiding versus accepting, predicted whether trauma symptoms followed.
A secondary analysis of 8,376 adolescents built two models to predict who benefits from school-based mindfulness training. Both found a subgroup — and the difference in outcomes was trivial.
The barriers parents named most often were cost and the absence of any nearby service — not stigma, which the field has spent years treating as the main obstacle.
Most psychedelic depression trials require patients to come off antidepressants first. This one deliberately didn't — and found intranasal 5-MeO-DMT tolerable and encouraging in patients who stayed on their SSRI.
A meta-regression of 169 trials coded 34 therapeutic ingredients. The most common — psychoeducation, relaxation, mindfulness — were not the ones that best predicted getting better.
The same drug sold as Zepbound and Mounjaro reduced cocaine-evoked dopamine release and blocked drug-associated place preference in rodents. It's animal data, not a human addiction treatment.
Up to eight sessions of psilocybin, layered onto a small randomized comparison against an active placebo, produced a 73% response rate by the end of treatment. The trial was designed to test safety first.
Six months out, 40.5% of smokers given one high dose of psilocybin plus therapy stayed abstinent, versus 10% on the nicotine patch plus the same therapy. The trial ran 82 people.
Compared with a diabetes drug from another class, GLP-1 agonists were followed by fewer new diagnoses of alcohol, opioid, nicotine and other substance use disorders.
A pooled analysis of five trials in treatment-resistant depression found sex differences in response. The largest ones appeared in both arms, which points at something other than the drug.
A 1,151-person Hong Kong trial used trained lay counsellors, not clinicians, and no internet. The effect sizes are small, and the authors are explicit that they do not know whether it travels.
In the RECOVER extension, about a fifth of participants were in remission at 24 months and some who had not responded at 12 months did later. The second year had no control group.
In 85 people with mild cognitive impairment, transient beta-band events recorded at rest differed years before diagnosis. The sample is small and the finding has not been replicated.
Seventy-three trials, and a familiar pattern: the benefit versus no treatment shrinks by a third when only the best-conducted studies are counted. The head-to-head comparisons rest on a few hundred people.
A population study covering all 15 million residents of Ontario found the rate of new ADHD medication prescriptions accelerated sharply after 2019 — and shifted decisively toward adults.
Bristol Myers Squibb found 'irregularities' in how a small number of ADEPT-2 sites executed the study, excluded their patients, and is enrolling more. Alzheimer's-related psychosis still has no approved treatment.
Semaglutide moved Alzheimer's biomarkers without moving the clinic. A separate liraglutide trial missed its primary endpoint on brain metabolism. Together they close a decade-old hypothesis — or at least narrow it.
A national registry analysis found the highest burden among 20-to-29-year-olds and a notable rise among girls aged 10 to 19, with rates concentrated in the Sierra region.
A meta-regression across 28 countries put pooled prevalence near 4-5% and standardised mortality ratios between 1.8 and 4.7, in a condition the Global Burden of Disease study does not count.
A within-individual analysis of 505,474 people in Finland and Sweden found lower psychiatric hospitalisation on clozapine in every disorder studied except borderline personality disorder, with no excess mortality signal.
Pooling 151 studies and 58,534 participants across 30 drugs, researchers ranked antidepressants by physiological side effects rather than efficacy. Some drugs raised cholesterol while reducing body weight.
Pooling 31 studies of 556 patients, cortical 5-HT2A binding was lower in unmedicated depression than in controls. The post-mortem studies mostly found nothing, and the tracers used shape the result.
Loneliness tracked with depression, perceived stress, psychotic-like experiences and suicidal ideation. A prevalence that high raises a question about what the measure is capturing.
Across 2.48 million survey responses in 159 countries, the share of people with no one to turn to rose 13.4% between 2009 and 2024 — all of it after 2019.
An Australian cohort of 3,044 children found no protective factor held across all ages and outcomes. School safety mattered at age 6-7; belonging mattered at 12-13.
In 270 patients whose voices persisted despite antipsychotics, the therapy beat matched clinical contact on hallucination severity by 2.26 points — significant, with a Cohen's d of 0.27.
The guidance is built on one negative instruction: do not stop abruptly in patients likely to be physically dependent. Hong Kong records show the sharpest prescribing rise in 18-to-25-year-olds.