Study archive · page 16

A sweep of randomised trials of digital mental health tools for older adults identified four features that tracked with success: ease of use, opportunities for social interaction, human support alongside the technology, and tailoring the tool to the person's needs. The review also flagged that future trials should follow people for up to 1 year afterwards to see whether benefits last.

⚠What it shows: Useful as a map rather than a measure: this review gathered randomised trials of mental health apps and programmes for older adults and pulled out what the successful ones had in common, but it did not pool the results into numbers. So nobody can say how much any single feature adds, or how long the benefits hold.Systematic review · Lancet Healthy Longev

Pooled results from 15 randomised trials in healthy people found that regular physical activity sustained over weeks improved working memory with a small effect, while a single acute bout of exercise produced no significant benefit.

⚠What it shows: Solid direction, small effect: results pooled from 15 trials where healthy people, from schoolchildren to retirees, were assigned at random to exercise or not. The trials used very different memory tests and activities, which is the main reason the single-session evidence stayed murky.Meta-analysis · Syst Rev

A meta-analysis of 18 randomised trials in 1354 healthy adults aged 60 or over found exercise training significantly reduced depressive symptoms, with benefits from both aerobic and resistance training, alongside improvements in quality of life measures, body mass, and upper and lower limb strength.

⚠What it shows: Solid pooled evidence: 18 trials in which 1354 healthy over-60s were assigned at random to exercise or comparison groups all point the same way, though the ideal amount and mix of exercise remains an open question.Meta-analysis · Biol Res Nurs

Pooling 13 randomised trials in 514 people with mild cognitive impairment, physical activity lowered the inflammatory markers TNF-α and CRP and raised the neuroprotective factors BDNF and IGF-1.

⚠What it shows: Real but indirect: results pooled from 13 trials in 514 people with mild cognitive impairment, showing modest shifts in blood markers. Nobody here measured whether memory or thinking actually improved, only the chemistry thought to sit behind it.Meta-analysis · Aging Clin Exp Res

Adding exercise to standard depression treatment produced moderately greater reductions in symptoms than standard treatment alone across 22 randomised trials, with the largest benefit in people with more severe depression.

⚠What it shows: Reliable direction, fuzzy size: results pooled from 22 trials all pointed the same way, but the trials disagreed with each other on how big the extra benefit is. The advantage held in the better-run studies and in people diagnosed with major depression.Meta-analysis · J Affect Disord

In a network meta-analysis of 69 randomised trials with 5,379 adults aged 65 and over, aerobic, resistance and mind-body exercise all eased depressive symptoms compared with controls, and no type beat the others head to head.

⚠What it shows: A dependable tie: results pooled from 69 trials in 5,379 adults aged 65 and over found every exercise type helped and none beat the others. The head-to-head estimates were wide, so a small edge for one style cannot be fully ruled out.Meta-analysis · F1000Res

A systematic review of seven randomised trials in 260 older adults found that resistance training three times a week, combined with at least 15 g of daily protein mostly from dairy, reduced blood markers that drive muscle loss and raised those that build it, alongside gains in strength, fat-free mass and muscle fibre size.

⚠What it shows: Promising but thin: a sweep of all the studies turned up just seven trials with 260 older adults between them, and they varied in length and design. The direction is consistent; the ideal amount of protein or training is not settled here.Systematic review · Biomolecules

Pooled results from 23 randomised trials in 1,134 older adults show elastic band resistance training improved every metabolic syndrome component measured: systolic blood pressure fell by 7.73 mmHg and fasting glucose by 9.02 mg/dL, while body fat, cholesterol and triglycerides also dropped compared with non-active controls.

⚠What it shows: Strong evidence for beginners over 65: results pooled from 23 trials in 1,134 people around age 70 who started out sedentary. Over nine in ten participants were women, so the picture for older men is much thinner.Meta-analysis · Arch Phys Med Rehabil

Pooled results from 13 randomised trials in 2,080 adults over 70 with sarcopenia show that more than 10 weeks of resistance training improves handgrip strength, muscle mass and chair-stand performance, but has no favourable effect on bone mineral density.

⚠What it shows: About as solid as exercise evidence gets: results pooled from 13 trials in which 2,080 adults over 70 with sarcopenia were assigned at random to train or not. The clear miss is bone, which did not improve, so the win here is muscle and strength.Meta-analysis · Geriatr Nurs

Pooling four randomised trials in 27,617 people, semaglutide lowered the risk of major cardiovascular events by 19% versus placebo, with clear benefits for cardiovascular death and non-fatal heart attack. It did not significantly reduce non-fatal stroke or hospital admissions for heart failure or unstable angina, and people on the drug were more likely to stop treatment.

⚠What it shows: About as solid as pooled evidence gets for a newer drug: four trials where 27,617 people were assigned at random to semaglutide or placebo, and the trials agreed closely on the heart benefit. They disagreed sharply on how many people quit the drug, so treat that part as rough.Meta-analysis · BMC Cardiovasc Disord

Across six randomised trials with 4036 participants, tirzepatide cut body weight versus placebo by 7.7 kg at the 5 mg dose and 11.8 kg at 15 mg, a 12.4% reduction at the top dose. Nausea, vomiting and diarrhoea were all markedly more common than on placebo.

⚠What it shows: Strong pooled evidence: six trials where 4036 people were given tirzepatide or placebo at random, running anywhere from 12 to 72 weeks. It says nothing about what happens to the weight once the injections stop.Meta-analysis · Int J Obes (Lond)

Pooled results from eight randomised trials show probiotic and prebiotic supplements do not change thyroid hormone levels, with TSH, free T4 and free T3 all unmoved. The only significant effect was a modest drop in thyroid antibodies in people with Graves' disease.

⚠What it shows: A solid null for healthy thyroids: results pooled from eight trials found no movement in any thyroid hormone measured. The antibody reduction applies to people with Graves' disease, a diagnosed autoimmune condition, not to healthy supplement users.Meta-analysis · PLoS One

In an Estonian cohort of 2,509 people with shotgun metagenomic gut sequencing linked to health records, past medication use was associated with the gut microbiome, with traces of antibiotics, psycholeptics, antidepressants, proton pump inhibitors and beta-blockers still detectable several years after use. In a subgroup of 328 people sampled twice, similar shifts appeared after starting or stopping treatment.

⚠What it shows: Suggestive rather than settled: researchers followed 2,509 people rather than testing drugs on them, so they cannot prove the medicines caused the differences, though repeat samples in 328 people showing shifts after starting or stopping treatment point that way.Study · mSystems

In trials pooled from postmenopausal women, hormone replacement therapy was associated with more lean body mass and less fat mass than placebo, while exercise training reduced fat mass without a clear effect on lean mass. A network comparison ranked oral estrogen and progestogen combined with exercise as the most effective option for reducing fat mass.

⚠What it shows: Reasonably solid on the main comparisons: results pooled from trials of anti-osteoporosis treatments in postmenopausal women, with differences that were real but fairly small. The claim that hormones plus exercise beats everything else comes from ranking treatments indirectly, not from a head-to-head trial, and exercise's effect on lean mass was too uncertain to call.Meta-analysis · Medicine (Baltimore)

Across 12 trials, exercise programmes during pregnancy and the postpartum period reduced postpartum depressive symptoms, with the largest benefit in mothers who already met criteria for depressive symptoms at the outset.

⚠What it shows: Fairly convincing: results pooled from 12 trials of exercise in pregnant and postpartum women, and the studies mostly pointed the same way. Some of the included trials did not assign women at random, so the true benefit may be somewhat smaller.Meta-analysis · Birth

Taking an isolated soluble fibre supplement for at least 12 weeks led to an average of 1.25 kg more weight loss in overweight and obese adults across 27 randomised trials, alongside significant drops in BMI, waist circumference, fasting insulin and insulin resistance.

⚠What it shows: About as solid as nutrition evidence gets: results pooled from 27 trials in 1,428 overweight or obese adults, with the reviewers rating the certainty as high. The average loss was still modest at just over a kilo, so it is a helper, not a weight-loss plan.Meta-analysis · Nutrients

Pooled results from five trials in 151 healthy adults found foam rolling or stick massage after damaging exercise made no significant difference to muscle soreness at any point from immediately afterwards to 96 hours, nor to range of motion, muscle swelling, or strength recovery.

⚠What it shows: Take it as a strong hint rather than a settled verdict: results pooled from only five trials covering 151 healthy adults, 136 of them men, and the quality was middling at best, so the authors themselves stop short of a definite conclusion.Meta-analysis · J Bodyw Mov Ther

Across 23 biological ageing clocks built from imaging, blood proteins and metabolites in UK Biobank participants aged 37 to 84, self-reported sleep duration showed a U-shaped relationship with biological age gaps, with the lowest gaps falling between 6.4 and 7.8 hours depending on organ and sex. Sleeping under 6 hours or over 8 hours, compared with 6 to 8 hours, was associated with higher risk of systemic diseases including depression and diabetes, and with all-cause mortality.

⚠What it shows: Strong data, soft measurement: a very large sample with 23 different ageing markers, but sleep was self-reported and people were followed rather than assigned sleep lengths, so illness driving short or long sleep cannot be ruled out.Study · Nature

Pooling 23 studies, training in hot conditions produced greater reductions in resting core temperature, exercising core temperature and exercising heart rate, plus larger increases in sweat rate and aerobic performance in the heat, than training in thermoneutral conditions; skin temperature during exercise was not significantly different. Training in thermoneutral conditions still improved exercising core temperature and heart rate compared with baseline.

⚠What it shows: Decent weight: results pooled from 23 studies that compared training in the heat with training in cool conditions, and the direction is consistent across most heat measures. The extra benefit was most pronounced in elite athletes and in longer programmes, so a recreational exerciser doing a few hot sessions may see considerably less.Review · J Sport Health Sci

A review of 13 randomised trials in 263 healthy adults suggests Rhodiola rosea taken acutely may improve endurance performance and lower perceived exertion, while chronic use may help anaerobic performance and muscle damage. Most included studies carried an unclear or high risk of bias.

⚠What it shows: Encouraging but far from settled: a sweep of all the studies on this found 13 trials covering 263 healthy adults, and most carried an unclear or high risk of bias. Doses and exercise tests varied so much that no single benefit is nailed down yet.Systematic review · Phytother Res

Pooled results from three double-blind trials in 231 people suggest oral probiotics modestly reduced inflammatory acne lesions compared with placebo, with no serious adverse events reported. The trials disagreed with each other, and the authors note future studies may plausibly find no meaningful effect.

⚠What it shows: Promising but thin: results pooled from just three trials in 231 people, and the trials disagreed with each other. The researchers' own analysis says future studies could plausibly find no meaningful effect at all.Meta-analysis · Medicina (Kaunas)

Researchers estimate that in Spain in 2020, 747 deaths from lung cancer and ischaemic heart disease were attributable to secondhand smoke, rising to 2242 deaths when chronic obstructive pulmonary disease and cerebrovascular disease were included. Of the attributable deaths, 60.9% occurred in non-smokers and 75.1% in men.

⚠What it shows: These are statistical estimates, not counted deaths: researchers combined national exposure surveys with risks pooled from earlier studies, so the totals depend on those inputs being right. The underlying link between secondhand smoke, lung cancer and heart disease is long established.Meta-analysis · Arch Bronconeumol

Pooled results from 21 randomised trials found that wearable devices promoting physical activity did not significantly lower systolic or diastolic blood pressure over 12 to 48 weeks in adults with hypertension, type 2 diabetes or overweight and obesity. Body weight, fasting blood glucose and HbA1c did not significantly improve either.

⚠What it shows: A fairly firm null as far as it goes: results pooled from 21 trials lasting 12 to 48 weeks, though the studies often disagreed with each other. The devices were mostly tested on their own, so combining them with coaching and support might still work differently.Meta-analysis · Hypertens Res

In 66 studies of resilience and drinking, with 24 of them pooled, higher personal resilience was associated with only slightly lower odds of alcohol-related outcomes (pooled odds ratio 0.91), and the studies disagreed with each other enormously. Emotion-regulation and coping components of resilience showed more consistent protective associations than competence or achievement-oriented traits.

⚠What it shows: Weak evidence for a popular idea: results pooled from 24 studies that surveyed or followed people rather than testing anything on them, so it cannot show that becoming more resilient makes anyone drink less. The studies also disagreed with each other almost completely, and two-fifths were done in school or college students, mostly in high-income countries.Review · Clin Psychol Rev