Study archive · page 34

Exercise programmes significantly reduced insomnia severity and eased sleep problems in menopausal women in a meta-analysis of randomised trials, with the clearest effects in women who already had sleep disorders. Overall sleep quality scores did not improve significantly.

⚠What it shows: Encouraging but uneven: results pooled from ten trials where women were assigned at random showed less insomnia but no clear change in overall sleep quality scores, and the underlying studies carried a moderate risk of bias. The review also could not tell which type or length of exercise works best.Systematic review · Front Med (Lausanne)

Pooled results from 21 randomised trials in 1252 university students with sleep problems found both traditional Chinese exercises and general aerobic exercise improved sleep, with aerobic exercise the more effective of the two. Effects were largest with sessions of 40 to 60 minutes, done 3 to 5 times a week.

⚠What it shows: Strong for students, unknown beyond them: results pooled from 21 trials covering 1252 university students, all with existing sleep problems. The trials disagreed with each other enough that the best session length and frequency should be read as a rough guide, not a prescription.Meta-analysis · Medicine (Baltimore)

A meta-analysis pooling 63 observational studies found a moderate correlation (r=0.38) between smartphone addiction and poor sleep quality in adults aged 18 to 25, alongside later sleep onset, shorter sleep and more daytime dysfunction.

⚠What it shows: Consistent but correlational: nearly all the underlying studies were cross-sectional and self-reported, so poor sleepers may simply reach for their phones more, not the reverse. The pooled link was moderate and results varied enormously between studies.Review · Int J Adolesc Med Health

A meta-analysis of 15 trials found that high-intensity exercise finishing 2 to 4 hours before bedtime did not disrupt nighttime sleep in healthy adults; the only significant change was a small 2.34% reduction in REM sleep.

⚠What it shows: A trustworthy pooled result built from small studies: 15 trials totalling just 194 healthy good sleepers aged 18 to 50. It says nothing about people who already sleep badly, or about sessions finishing very close to lights out.Meta-analysis · Sleep Med Rev

A systematic review of 13 randomised trials found that regular resistance training improves all aspects of sleep, with the greatest benefit for sleep quality, and also improves anxiety and depression.

⚠What it shows: A credible review of a thin evidence base: just 13 trials, and the sleep benefit shrank when lifting was combined with aerobic work rather than done on its own. What a single session does to that night's sleep remains poorly studied and inconsistent.Systematic review · Sleep Med Rev

Strength training improves the nervous system's ability to switch muscle on fully in older adults, with voluntary activation rising by 8.8% in the calf muscles and 1.8% in the knee extensors across pooled trials. Gains were largest in people who started with the poorest activation.

⚠What it shows: Solid but narrow: results pooled from 17 studies of lower-body training in older adults, and the effect on the opposing muscles that brake a movement was inconsistent, so that piece of the puzzle is still open.Meta-analysis · Exp Gerontol

Training a muscle once, twice or three or more times per week produces similar growth when total weekly volume is the same, according to a meta-analysis of 25 studies. Higher frequency only helped when it brought extra total work with it.

⚠What it shows: Strong evidence by fitness research standards: results pooled from 25 studies, the finding held for trained lifters and for upper and lower body alike. The one nuance is that frequency did help when it was not matched for total work, meaning extra sessions mostly matter as a vehicle for extra volume.Meta-analysis · J Sports Sci

Higher-volume resistance training beat lower volumes for building both upper and lower body strength in adults aged 60 to 83, in a meta-analysis of 31 studies with 1,744 participants. The advantage was most pronounced after 12 weeks of training.

⚠What it shows: A consistent pattern rather than a precise prescription: results pooled from 31 studies of 1,744 independent older adults, and it does not settle exactly how many sets count as enough. The functional fitness benefit was a tendency, not a firm result.Meta-analysis · Arch Gerontol Geriatr

A network meta-analysis of 75 randomised trials in 4048 women with overweight and obesity found each exercise type topped a different inflammatory marker: resistance training ranked best for lowering C-reactive protein, aerobic exercise for TNF-α, combined training for interleukin-6 and leptin, and HIIT for raising adiponectin.

⚠What it shows: Strong on breadth, weaker on head-to-head certainty: the rankings pool 75 trials with 4048 participants, but network meta-analysis compares exercise types partly indirectly, across studies rather than within them. The winner for one marker was not the winner for another, so no single verdict emerges.Meta-analysis · J Sci Med Sport

Exercise improved self-reported erectile function in men diagnosed with erectile dysfunction, with a pooled gain of 3.85 points on the International Index of Erectile Function across 7 trials and 478 participants. Aerobic exercise at moderate to vigorous intensity showed the clearest benefit.

⚠What it shows: Worth taking seriously: results pooled from 7 trials in 478 men with diagnosed erectile dysfunction, and the benefit held across programmes lasting 8 weeks to 2 years. Nobody could be blinded to whether they were exercising and the score is self-reported, which can flatter the result.Meta-analysis · Br J Sports Med

In a cohort of 17,088 UK adults with activity trackers, meeting the 150 minutes a week guideline was linked to a modest 8%-9% lower cardiovascular disease risk; a reduction beyond 30% required roughly 560-610 minutes a week.

⚠What it shows: Worth weight but not proof of cause: this is an observational study of 17,088 UK adults who wore activity trackers, with genetic analyses pointing the same way for fitness. Activity was measured once, and Biobank volunteers tend to be healthier than the general population.Study · Br J Sports Med

A meta-analysis of 21 randomised trials found that caffeinated energy drinks raised systolic blood pressure by 9.02 mmHg and heart rate by 2.86 bpm during exercise in healthy adults, while diastolic blood pressure was unchanged.

⚠What it shows: Solid pooled evidence, but a narrow window on it: 21 randomised trials covering just 296 healthy adults, mostly men, during aerobic exercise only. The data were too thin to say anything about heart rhythm effects, which is the question many readers will actually have.Meta-analysis · Appl Physiol Nutr Metab

A systematic review of trials published between 2011 and 2024 finds that leucine-enriched whey protein combined with resistance training improves lean muscle mass, strength and physical performance in older adults with sarcopenia. Optimal dosing and long-term sustainability remain open questions.

⚠What it shows: Encouraging but imprecise: a sweep of all the studies on this from 2011 to 2024 points the same way, but the review reports no pooled numbers, and the authors themselves flag that ideal doses and long-term effects are still unknown.Systematic review · J Cachexia Sarcopenia Muscle

In a systematic review of 16 controlled trials in adult men, N-acetylcysteine (NAC) supplementation improved exercise performance, antioxidant capacity and glutathione status with no serious side effects reported, but showed no clear benefit for blood markers, inflammation or measures of muscle behaviour.

⚠What it shows: Promising but far from settled: a sweep of all the studies turned up 16 controlled trials in adult men, most reporting benefits, but the results were never pooled into a single number and the authors themselves call for more research. Women were not studied, so nothing here applies to them yet.Systematic review · Nutrients

A single 20 mg dose of cortisol markedly impaired path integration, the ability to keep track of your own position while moving, in 39 healthy men tested under both cortisol and placebo, and diminished grid-like activity in the entorhinal cortex on brain scans.

⚠What it shows: Interesting but early: 39 healthy men were each tested under both cortisol and a placebo in a lab, so it says nothing yet about women, real-world stress or repeated exposure. The dose came from a pill, not from stress itself.Study · PLoS Biol

In a pre-registered study of 251 UK adults rating 20 real-life challenges, eustress, the positive experience of stress, behaved as a measurable trait: nine processes such as successful outcomes, fulfilment and positive affect explained a median 83% of the variance in how positive a challenge felt, and it was unrelated to trait distress.

⚠What it shows: Useful groundwork rather than a life hack: 251 UK adults rated challenging situations in one pre-registered survey, so it maps what good stress looks like without testing how to create it. Everything rests on self-report.Study · Soc Indic Res

The triglyceride-glucose index, a cheap marker built from routine blood tests, was linked with roughly double the cardiovascular risk in a Chinese cohort, yet adding it to the Framingham Risk Score barely improved prediction, and pooled results across five cohorts found a negligible gain.

⚠What it shows: A useful corrective rather than a final word: one Chinese study following adults aged 45 to 80, plus results pooled from five similar studies, none of which tested the marker in a trial. The authors themselves rated the certainty on the prediction question as very low.Study · Front Cardiovasc Med

Across 97 randomised trials in adults with overweight and obesity, adding exercise to a low-calorie diet improved blood glucose, insulin, insulin resistance, triglycerides, HDL cholesterol and blood pressure more than diet alone, with the benefits tied mainly to aerobic exercise, especially supervised high-intensity cycling, rather than resistance training.

⚠What it shows: Broad but not bulletproof: results pooled from 97 trials in adults with overweight or obesity, yet the authors rate the evidence as only moderate to low certainty. The clearest gains came from supervised high-intensity programmes, so casual unsupervised exercise may not deliver the same result.Meta-analysis · Nutr Rev

Across 12 population-based studies of 676,658 people, links between thyroid problems and anxiety were small and mostly not statistically significant. Only an overactive thyroid showed a modest significant association.

⚠What it shows: Take the uncertainty seriously: this pools 12 population studies of 676,658 people, the most representative evidence yet, but the studies disagreed with each other, so anything from a modest link to no link at all remains possible.Meta-analysis · PeerJ

In a trial of 5246 men aged 45 to 80 with low testosterone and high cardiovascular risk, testosterone gel did not increase major cardiac events compared with placebo, with events in 7.0% versus 7.3%. Atrial fibrillation, acute kidney injury and pulmonary embolism were more common with testosterone.

⚠What it shows: About as solid as single-trial safety evidence gets: 5246 men with low testosterone and existing heart risk were assigned at random to testosterone gel or placebo and followed for nearly three years on average. It shows no extra heart attacks or strokes, not that testosterone helps, and clots, heart rhythm problems and kidney injury were more common on treatment.RCT · N Engl J Med

Adding three daily servings of full-fat dairy for 12 weeks left body weight, fat mass and fasting cholesterol unchanged in 74 adults with overweight or obesity, even with no calorie restriction. The group eating it freely also saw small falls in hip circumference and systolic blood pressure.

⚠What it shows: Reassurance rather than a green light: one 12-week trial in 74 adults with overweight, showing full-fat dairy did no harm rather than that it helps. The dairy groups also ate more protein and calcium, which muddies what caused the small benefits.RCT · J Nutr

Resistance training was associated with 21% lower all-cause mortality on its own, and 40% lower when combined with aerobic exercise, in a meta-analysis of 11 studies totalling 370,256 people followed for an average of 8.85 years.

⚠What it shows: Compelling but not causal: 10 of the 11 pooled studies followed people rather than assigning them to lift, so healthier people may simply be more likely to train. The link with cancer deaths was not significant and cardiovascular mortality was only borderline.Meta-analysis · Eur J Prev Cardiol

Oral hydrolysed collagen taken for 90 days reduced wrinkles and improved skin hydration and elasticity compared with placebo in a meta-analysis of 19 randomised, double-blind trials with 1,125 participants aged 20 to 70, 95% of them women.

⚠What it shows: Stronger than most supplement claims: results pooled from 19 placebo-controlled trials in 1,125 people, where neither participants nor researchers knew who got the real thing. Almost all participants were women aged 20 to 70, so the evidence says little about men.Meta-analysis · Int J Dermatol

Across 12 randomised trials in 482 menopausal and postmenopausal women, resistance training reduced C-reactive protein, a blood marker of inflammation, by a mean of 0.47 mg/dL, though the reviewers rated the evidence as very low quality.

⚠What it shows: Treat this as a promising signal rather than a settled result: 12 randomised trials in 482 women pointed the same way, but the reviewers themselves graded the evidence very low quality, citing study flaws, inconsistency and possible publication bias. They state plainly that the true effect could be substantially different.Meta-analysis · Menopause