Study archive · page 18

In head-to-head trials in adults with type 2 diabetes, semaglutide lowered HbA1c by an extra 0.44% and body weight by an extra 2.53 kg compared with other GLP-1 drugs, but it also caused significantly more gastrointestinal side effects and more people stopping treatment.

⚠What it shows: Strong for what it measures: results pooled from 5 head-to-head trials in 3760 people with type 2 diabetes, so the comparison is direct rather than inferred. It tracked blood markers and weight rather than heart attacks or deaths, and the extra benefit came with more people quitting the drug.Meta-analysis · J Diabetes Complications

Pooled across 11 randomised trials, workplace health programmes for men cut body weight by an average of 0.28 kg within 3 months and 1.38 kg beyond 3 months, changes that were not statistically significant despite improvements in snacking, sugary drink intake and physical activity.

⚠What it shows: Suggestive at best: results pooled from 11 trials in working men pointed the right way, but none of the weight changes reached statistical significance. Most trials were short, so nobody knows whether these programmes deliver over years.Meta-analysis · Health Educ Behav

In pooled data from 68 randomised trials with 207 200 participants, only semaglutide was associated with a higher incidence of colorectal tumours, and only in its high-dose injectable form at 2.4 mg per week; other GLP-1 receptor agonists and all SGLT2 inhibitors showed no such association.

⚠What it shows: An early warning light, not a verdict: pooled data from 68 trials with 207 200 people can surface rare events, and the authors themselves call this a signal needing validation. People with obesity already carry a higher baseline colorectal risk, which complicates interpretation.Meta-analysis · Int J Surg

Pooled results from 91 studies of 4431 healthy adults show a month of dawn-to-sunset fasting during Ramadan produced small favourable shifts in heart-risk markers, most clearly a rise in HDL cholesterol and a fall in LDL cholesterol. The authors describe any protection as short-term and transient.

⚠What it shows: A big evidence base but small effects: pooled from 91 studies of 4431 healthy adults across 23 countries. Most changes were slight, and the authors themselves call the protection short-term and transient, so nothing here shows the benefit outlasts the fasting month.Meta-analysis · Nutr Metab Cardiovasc Dis

In pooled results from 11 published and four unpublished randomised trials, the GLP-1 and glucagon dual agonists mazdutide and cotadutide cut body weight by 4.16% and HbA1c by 0.63% versus placebo in people with type 2 diabetes, obesity or both. Serious adverse events did not rise, but the odds of vomiting were 6.05 times higher.

⚠What it shows: A solid pooled picture: results combined from 11 published and four unpublished trials against placebo, so the benefits look real, though four of the datasets have not passed peer review and everyday side effects such as vomiting were far more common on the drugs.Meta-analysis · Endocrine

A sweep of 68 studies found that people with type 2 diabetes who fast during Ramadan have a lower risk of hypoglycaemic events on non-sulfonylurea medications than on sulfonylureas, with risk also shaped by diet, fasting length, sex, season, time since diagnosis and pre-fasting education.

⚠What it shows: Broad but not boiled down to one number: a sweep of 68 studies, 35 trials where people were assigned at random and 33 that followed people through the month, all in adults with type 2 diabetes. It maps who is most at risk rather than pooling a single figure, so exact risk sizes vary between studies.Systematic review · Front Endocrinol (Lausanne)

Twice-daily peppermint oil supplementation, 100 μL per day, lowered systolic blood pressure by 8.48 mmHg relative to placebo after 20 days in a randomised trial of 40 adults with pre-hypertension and stage 1 hypertension. Readings fell from 130.05 to 121.97 mmHg in the peppermint group while the placebo group stayed flat.

⚠What it shows: Encouraging but far from settled: one 20-day trial in just 40 people with mildly raised blood pressure, run by a single research group. It needs repeating in bigger groups, and for longer, before peppermint earns a place alongside proven blood pressure care.RCT · PLoS One

In a snapshot study of 54 older adults with mild cognitive impairment and disturbed sleep, those taking 5000 IU or more of vitamin D daily scored higher on the MoCA cognitive test than people taking none. Lower doses showed no significant link, and D3 users did no better than D2 users.

⚠What it shows: Treat this as a lead, not an answer: a one-off snapshot of 54 people whose supplement use was self-reported, so it cannot show the vitamin caused the better scores. Being overweight also tracked with higher scores here, a hint that things other than the pill may be driving the pattern.Study · Sleep Med

Across 73 trials in 1061 healthy adults, dietary nitrate supplementation improved power output by 4.6 watts, time to exhaustion by 25.3 seconds and distance covered by 163.7 metres versus placebo. It did not improve time-trial performance, perceived exertion, VO2max or blood lactate.

⚠What it shows: Strong evidence overall: results pooled from 73 trials covering 1061 healthy adults, graded with a standard confidence framework. The catch is that the gains were modest and did not show up in time-trial performance, the test closest to actual racing.Meta-analysis · J Int Soc Sports Nutr

Pooled across 22 randomised trials in 1551 people, probiotics, prebiotics and fermented foods produced no significant improvement in global cognition or in any individual cognitive domain, despite 14 individual studies reporting benefits.

⚠What it shows: A trustworthy null: results pooled from 22 trials in 1551 people, most of them well run, found no cognitive benefit anywhere. The trials were small and short, so a subtle long-term effect cannot be fully ruled out.Meta-analysis · Neurosci Biobehav Rev

In people with type 2 diabetes, dietary fibre increased the beneficial gut bacterium Bifidobacterium and lowered circulating lipopolysaccharide, total cholesterol and BMI compared with control diets, but left triglycerides, HDL, LDL and most inflammatory markers unchanged.

⚠What it shows: Modest and mixed: results pooled from up to 11 trials in people with type 2 diabetes, with some outcomes resting on barely more than a hundred participants. Short trials and very different fibre types and amounts make the true size of the benefit uncertain.Meta-analysis · Nutrients

Higher relative abundance of the gut species Roseburia inulinivorans, but not of other Roseburia species, was associated with stronger handgrip, leg press and bench press in adults, and the species was less abundant in older adults than younger ones. Supplementing it in antibiotic-treated mice increased forelimb grip strength, alongside larger muscle fibres and a shift from type I to type II fibres.

⚠What it shows: Early and split evidence: in people this is only an association between one gut species and strength scores in groups of younger and older adults, and the test of cause was done in mice given antibiotics. Nobody has yet given this bacterium to humans to see whether strength changes.Study · Gut

In women with PCOS, lower glycaemic index diets reduced insulin resistance, total and LDL cholesterol, triglycerides, waist circumference and total testosterone compared with higher-GI diets, without changing fasting glucose, HDL cholesterol or body weight.

⚠What it shows: Encouraging but not settled: results pooled from 10 trials in 403 women with PCOS, and the underlying trials disagreed with each other on the main insulin measure. Effects on ovulation, fertility and heart disease were never tested.Systematic review · Adv Nutr

Pooled results from 26 randomised trials in 2,170 postmenopausal women show exercise significantly reduced depressive symptoms, with mind-body exercise the most likely to be the most effective type, ahead of aerobic and multicomponent exercise. Anxiety also improved.

⚠What it shows: Encouraging but not settled: results pooled from 26 trials in 2,170 postmenopausal women, yet the review's own grading rated confidence in many comparisons as very low, so the ranking of exercise types could still shift as better trials arrive.Meta-analysis · BMC Public Health

Across 91 randomised trials including 89,558 women, high-dose tirzepatide (15mg/week) was the only GLP-1 or SGLT2 regimen associated with increased overall gynaecologic tumour risk, an absolute risk difference of 0.57%, and both high and low tirzepatide doses were linked to intra-uterine tumours. The authors caution the association may not be causal.

⚠What it shows: A signal, not a verdict: pooled from 91 drug trials covering 89,558 women, but tumour cases were rare, the risk estimates carry wide uncertainty, and the authors themselves say more data are needed to rule out chance or bias.Meta-analysis · J Hematol Oncol

Non-drug approaches improved new mothers' self-reported sleep in pooled trial data, with massage and exercise showing the largest effects. Infants' night-time sleep also improved slightly, but maternal depression did not shift.

⚠What it shows: Reasonably solid but resting on self-report: results pooled from 15 studies of new mothers, with sleep judged by the mothers themselves rather than measured objectively. Whether the gains last, or show up on sleep recordings, is still unknown.Meta-analysis · Sleep Med Rev

Across 34 randomised trials in 3078 adults aged 60 and over, cognitive behavioural therapy for insomnia, especially when combined with positive mood strategies, ranked as the most effective of 21 drug-free approaches to improving sleep quality.

⚠What it shows: Useful rankings, low certainty: results pooled from 34 trials in 3078 older adults compared 21 approaches indirectly, and the authors rate the overall evidence as low certainty. Which option came out best also shifted with how bad sleep was to begin with.Meta-analysis · BMC Med

Pooled trial data suggest the intensity of a strength session is not associated with how many calories it burns; once weaker studies were accounted for, the apparent effect of intensity on energy expenditure disappeared. Traditional straight-set training did show higher energy expenditure than circuit training in the pooled comparison.

⚠What it shows: Strong as pooled evidence goes, but the underlying trials disagreed with each other a lot. Once the analysis accounted for weaker studies, the apparent link between intensity and calorie burn vanished entirely.Meta-analysis · Clin Physiol Funct Imaging

Across 54 randomised trials in 2,836 adults aged 60 and over, exercise was associated with improved inflammatory markers, and aerobic exercise showed the most consistent signal for reducing C-reactive protein, with an average standardised reduction of 0.83.

⚠What it shows: Solid for the big picture, shakier for the league table: results pooled from 54 trials in 2,836 adults aged 60 and over. Some rankings, especially for a second inflammation marker, collapsed once one extreme study was removed, so treat the ordering of exercise types with caution.Review · Front Med (Lausanne)

Pooling 15 studies of 339 physically active adults, listening to music during aerobic exercise was most consistently linked to lower perceived exertion. It was also associated with better fixed-duration performance and lower blood lactate, while heart rate was not significantly affected.

⚠What it shows: Solid for the effort finding, shakier elsewhere: results pooled from 15 studies covering just 339 active adults. The endurance studies disagreed with each other so much that the apparent boost to time to exhaustion deserves very little weight.Systematic review · Front Sports Act Living

Pooling 42 randomised trials in 4,740 frail older adults, physical activity produced small to moderate gains in global cognition, attention and executive function, and moderate to large gains in psychological well-being and quality of life; memory did not improve.

⚠What it shows: Strong for the headline effects: results pooled from 42 trials in 4,740 older adults with physical or cognitive frailty. The programmes varied a lot between trials, and memory, the outcome people care most about, did not budge.Meta-analysis · J Phys Act Health

Pooled data from 16 studies of 664 healthy adults show Ramadan intermittent fasting significantly raised ghrelin, the hormone that drives hunger, while leptin, insulin and gastrin showed no significant change.

⚠What it shows: Reasonably solid on the headline but thin underneath: results pooled from 16 studies of 664 healthy adults, yet only three of those studies measured ghrelin. The leptin studies disagreed with each other, so that null is less settled than it looks.Meta-analysis · Clin Nutr

A sweep of the research found that people with higher historical air pollution exposure tended to score worse on tests of cognitive function, though the two studies tracking decline over time disagreed with each other.

⚠What it shows: Suggestive at best: a sweep of all the studies on this found only eight worth including, most relying on indirect estimates of pollution exposure and unclear cognitive testing. It can show that pollution and worse scores travel together, not that one causes the other.Systematic review · Age Ageing

A pooled analysis of 46 randomised trials in 3463 people with knee osteoarthritis ranked high-speed resistance training as the most effective type for easing pain and stiffness and improving function. For pain and function, moderate loads of 43 to 47% of one-rep max appeared optimal.

⚠What it shows: Strong as exercise evidence goes: results pooled from 46 trials in 3463 people with knee osteoarthritis. The trials varied a lot in how they ran the training, so the precise dose numbers are best read as ballpark targets rather than rules.Meta-analysis · BMC Musculoskelet Disord