Study archive · page 27

Combining aerobic and resistance training ranked as the most effective exercise approach for improving cardiometabolic health in adults with overweight or obesity, across a network analysis of 81 randomised trials.

⚠What it shows: A robust ranking built from results pooled across 81 trials in 4331 adults with overweight or obesity. Some comparisons between exercise types are indirect, and the effects appeared to differ between men and women.Meta-analysis · Circ Cardiovasc Qual Outcomes

Pooled data from 67 studies of 2058 people show muscle growth and strength both climb as weekly set volume rises, but with diminishing returns, and the flattening is far steeper for strength. Training a muscle more often reliably helped strength, yet its effect on muscle size was compatible with negligible.

⚠What it shows: Strong evidence: results pooled from 67 studies of 2058 mostly young lifters, around 79% of them men. Averages across studies cannot tell you exactly where your own point of diminishing returns sits.Meta-analysis · Sports Med

In pooled results from 22 randomised trials of 959 older adults with sarcopenia, resistance training produced large gains in handgrip and knee extension strength, small gains in muscle mass, and no reduction in body fat percentage.

⚠What it shows: Reasonably solid: results pooled from 22 trials in 959 older adults diagnosed with sarcopenia, so the findings speak to that group rather than healthy younger lifters. Programmes varied widely, which blurs exactly which routine works best.Meta-analysis · Arch Gerontol Geriatr

Pooled results from nine trials show unilateral and bilateral resistance training build muscle to a similar degree. Strength gains were specific: bilateral training improved two-limb strength more, while unilateral training improved single-limb strength more.

⚠What it shows: Reasonable but thin: results pooled from just nine trials, most carrying a moderate risk of bias, so the no-difference verdict on muscle growth could shift as more data arrive. The strength finding is on firmer ground because gains consistently tracked the way people trained.Meta-analysis · Sports Med

Pooled data from 17 controlled studies in 262 endurance cyclists show heavy strength training improved cycling performance, pedalling efficiency and anaerobic power, with no change in VO2max or metabolic thresholds.

⚠What it shows: Promising, but the reviewers themselves rate the evidence as low certainty: results pooled from 17 controlled studies of 262 cyclists, only 60 of them women, with programmes lasting 5 to 25 weeks. Not enough yet to say exactly how the lifting should be programmed.Meta-analysis · Eur J Appl Physiol

GLP-1 receptor agonist drugs cut body weight by an average of 4.57 kg and trimmed 4.55 cm off the waist compared with placebo, across a pooled analysis of 47 randomised trials in adults with overweight or obesity.

⚠What it shows: Strong evidence: results pooled from 47 trials in 23,244 adults with overweight or obesity. The trials varied widely in drugs, doses and populations, so the averages hide big differences from one drug and one person to the next.Meta-analysis · Diabetes Care

In adults with overweight or obesity but no diabetes, tirzepatide produced up to 17.8% weight loss, semaglutide up to 13.9%, and the premarket drug retatrutide up to 22.1% across a review of 26 randomised trials. Side effects were common but mostly gastrointestinal, and rarely forced people to stop treatment.

⚠What it shows: A thorough sweep of 26 placebo-controlled trials in 15,491 people without diabetes, but the trials differed too much to pool into one number. None compared the drugs head to head, so the league table rests on separate trials in different populations.Systematic review · Ann Intern Med

A single day of continuous glucose monitor data may reveal hidden diabetes risk in people with prediabetes: those with a High Blood Glucose Index above 0.174 had a 5.39- to 6.13-fold higher risk of progressing to type 2 diabetes, even when their standard glucose readings sat within recommended ranges. The metric behaved consistently across three different monitor brands.

⚠What it shows: Promising but early: the finding comes from 365 glucose monitor records across four separate groups of people, and so few actually progressed to diabetes that the 0.174 cut-off still needs confirming in bigger, independent groups before anyone acts on it.Study · Diabet Med

In adults with type 2 diabetes, every 30 minutes per week of supervised aerobic exercise cut HbA1c by 0.22 percentage points, with the benefit climbing to a 0.96 point drop at 100 minutes per week. Beyond 100 minutes the curve flattened, so extra weekly cardio bought no further HbA1c reduction.

⚠What it shows: About as strong as exercise evidence gets for the headline effect, which the authors rated high certainty: results pooled from 26 trials in 1253 adults with type 2 diabetes. The medication and side-effect numbers rest on far fewer trials and are much softer.Meta-analysis · Sports Med

Across 19 pooled studies following 25,376 people, mostly in Asia, each 1-unit rise in the triglyceride-glucose (TyG) index, a low-cost blood marker of insulin resistance, was associated with a 114% higher risk of atrial fibrillation.

⚠What it shows: Treat this as a strong signal rather than a verdict: results pooled from 19 studies that followed 25,376 people rather than testing anything on them, so it cannot prove the marker causes the rhythm problem. Nearly all the data come from Asian populations and the studies disagreed on the size of the link, so the exact figures travel with caution.Review · Front Cardiovasc Med

In pooled results from 16 randomised trials totalling 810 people, grape seed extract lowered systolic blood pressure by about 6 mmHg and diastolic by nearly 3 mmHg, with the biggest effects in younger, heavier participants and those with metabolic syndrome.

⚠What it shows: Promising but thin: results pooled from 16 trials covering just 810 people in total. The authors themselves call for a large, long-term trial in people with high blood pressure before anyone counts on it.Meta-analysis · Medicine (Baltimore)

Pooled results from 24 studies in women show caffeine before resistance exercise slightly improves maximal strength and may boost muscular endurance in some protocols, but the certainty of evidence is low or very low and results vary widely.

⚠What it shows: Treat it as a starting point: a sweep of 24 studies in women found small average benefits, but the reviewers rate their own certainty as low or very low. Results swung so much between protocols that no dose, timing or population advice can be drawn from it.Systematic review · Front Physiol

Across 18 pooled trials in 1023 people with diabetes, prediabetes or gestational diabetes, zinc supplements improved insulin resistance and lowered inflammation and oxidative stress markers. But LDL cholesterol rose slightly, by a mean of 3.46, a trade-off whose clinical relevance remains uncertain.

⚠What it shows: Reasonably solid on direction: results pooled from 18 trials in 1023 people with diabetes, prediabetes or gestational diabetes. The wrinkle is the mixed cholesterol picture, and nobody yet knows whether the small LDL rise matters in practice.Meta-analysis · Endocrinol Diabetes Metab

Pooled results from 24 calorie-matched trials in 1063 adults show higher-protein low-fat diets produced 0.79 kg more weight loss and 0.87 kg more fat loss than standard-protein versions, while better preserving fat-free mass and resting energy expenditure.

⚠What it shows: About as solid as diet science gets: results pooled from 24 trials in 1063 adults, all prescribed the same calories and fat. The average diet ran only around 12 weeks, so whether the protein edge survives long term is unknown.Meta-analysis · Am J Clin Nutr

Pooled results from 26 randomised trials in 1461 adults found fermented foods lowered the inflammatory marker TNF-α but did not change CRP or IL-6.

⚠What it shows: Take it lightly: results pooled from 26 trials in 1461 generally healthy adults, but the trials disagreed with each other so strongly that even the average TNF-α drop is hard to trust. The two most commonly cited inflammation markers did not move at all.Meta-analysis · Clin Nutr ESPEN

In older adults with insomnia, muscle-strengthening exercise improved sleep quality more than aerobic or combined exercise, lowering the global sleep quality index by 5.75 points across a network analysis of 25 trials.

⚠What it shows: A credible ranking but not the final word: results pooled from 25 trials in 2170 older adults with insomnia, and the trials often disagreed with each other sharply. Many of the comparisons between exercise types are indirect rather than head to head.Meta-analysis · Fam Med Community Health

Pooled results from 18 trials in 1214 adults with sleep disorders found exercise improved sleep efficiency, cut time awake after falling asleep and extended deep slow-wave sleep. Moderate-intensity aerobic exercise done 3 times a week for 45-60 minutes over 8-12 weeks ranked best for efficiency and night waking, while training 4 or more times a week did more for deep sleep.

⚠What it shows: Credible for the overall effect: results pooled from 18 trials in 1214 adults with diagnosed sleep disorders. The ranking of which exercise style wins rests on indirect comparisons between studies that differed from each other, so treat the 'best modality' verdict loosely.Meta-analysis · Sleep Med

In a meta-analysis of 116 trials with 6880 adults with overweight or obesity, every extra 30 minutes per week of aerobic exercise was linked to about 0.52 kg less body weight, 0.56 cm off the waist and 0.37% less body fat. Clinically important reductions appeared only beyond 150 minutes per week at moderate intensity or higher.

⚠What it shows: Solid evidence: results pooled from 116 trials of 6880 adults with overweight or obesity, so the per-half-hour effect is real but modest. The programmes were supervised, which means unsupervised real-world exercise may deliver less.Meta-analysis · JAMA Netw Open

In adults aged 65 and over, pooled results from 15 randomised trials in 480 people show both training styles lift peak oxygen uptake, but intervals win. HIIT improved VO2peak by 4.61 versus controls and beat continuous endurance training head to head by 3.76, while endurance training alone added 1.35 over controls.

⚠What it shows: Fairly convincing: results pooled from 15 trials where people were assigned at random, covering 480 adults aged 65 or older, with none judged at high risk of bias. The authors still want further trials before treating the size of HIIT's edge as settled.Meta-analysis · Int J Clin Pract

In older adults worried about their memory, lifestyle programmes worked best when they built capability, opportunity and motivation together; social support plus sustained nutrition counselling improved diet, and structured, tailored training improved physical activity. Evidence that behaviour change itself improved cognition remained low certainty.

⚠What it shows: Take the behaviour change finding seriously and the brain finding lightly: a sweep of 23 trials found moderate-certainty improvements in diet and activity, but only low-certainty evidence tying those changes to better cognition. Hardly any trials recorded ethnicity or income, so nobody knows whether this works for underserved groups.Review · Ageing Res Rev

Across 11 randomised placebo-controlled trials in 805 people, oral collagen supplements improved skin elasticity, hydration and dermal collagen density, with no adverse events reported. Most trials used collagen hydrolysate at 2.5 g to 10 g a day for 8 to 24 weeks.

⚠What it shows: Promising but early: a sweep of all the placebo-controlled trials found 11, totalling 805 people, and they varied in product, dose and duration. The reviewers themselves call the results preliminary, and the best dosing regimen is still unknown.Systematic review · J Drugs Dermatol

A Cochrane review of 249 trials found exercise reduced chronic low back pain by a clinically important margin compared with no treatment, usual care or placebo, and is probably more effective than education alone or non-exercise physical therapy. Improvements in day-to-day function were smaller and fell short of the threshold for clinical importance.

⚠What it shows: About as trustworthy as back pain evidence gets: a Cochrane sweep of 249 trials with moderate certainty in the main result. Pain relief cleared the bar for a difference patients would feel, but gains in daily function were small, and most trials could not hide from participants which treatment they received.Meta-analysis · Cochrane Database Syst Rev

In pooled trials covering 10,708 people from the general community, taking vitamin C daily did not prevent colds. Yet in marathon runners, skiers and soldiers under severe physical stress it roughly halved cold risk, and regular use shortened colds by 8% in adults and 14% in children.

⚠What it shows: Definitive for the average person: pooled trials in over ten thousand community members show vitamin C does not stop you catching colds. The halved risk applied only to 598 people under extreme exertion, and taking it once symptoms start showed no consistent effect.Meta-analysis · Cochrane Database Syst Rev

Pooling data from 78 studies, exercise performance was at most trivially reduced during the early follicular phase of the menstrual cycle, with no meaningful differences across other phases. The authors conclude the evidence is too weak and variable to support general cycle-based training guidelines, recommending a personalised approach instead.

⚠What it shows: Treat this as the best current estimate rather than the final word: results pooled from 78 studies found only a trivial dip, but the reviewers rated the overall evidence quality as low. Individual responses vary enough that no blanket rule fits everyone.Meta-analysis · Sports Med