Study archive · page 29

A meta-analysis of 20 randomised trials in women found acute caffeine gave a small average performance boost (Hedges' g = 0.37), with positive estimates in every menstrual-cycle phase and in contraceptive users, and no clear evidence that phase changes the effect. Certainty was low to very low.

⚠What it shows: Systematic review and meta-analysis of 20 randomised trials with 144 effect sizes in women; risk-of-bias concerns, sparse subgroup data, inconsistent verification of menstrual phase and low-to-very-low GRADE certainty mean subgroup conclusions are hypothesis-generating only.Systematic review · Front Nutr

Across 13 randomised placebo-controlled trials in rugby players, acute caffeine produced small improvements in sprint performance and lower perceived exertion, plus a large but highly heterogeneous effect on passing accuracy. Jumping, strength, metabolic and hormonal markers showed no significant change.

⚠What it shows: Systematic review and meta-analysis of 15 studies (13 in the pooled analysis) in rugby players, but GRADE certainty was low to very low, and the passing-accuracy result showed high heterogeneity and possible small-study effects.Systematic review · Front Nutr

In the 573-person clinic subcohort of the COSMOS trial, a daily multivitamin-mineral supplement produced a more favourable two-year change in episodic memory than placebo (0.12 SD units) but no clear benefit for executive function or attention. Pooling three COSMOS cognitive substudies, the multivitamin improved global cognition by 0.07 SD units, an effect the authors equate to reducing cognitive ageing by 2 years.

⚠What it shows: Meta-analysis of three cognitive substudies (n=573, 2158 and 2472) nested within the single COSMOS trial of adults aged 60 and over, so it is not independent replication; effects were small in size, tests differed between substudies, and no effect was seen on executive function or attention in the clinic subcohort.Meta-analysis · Am J Clin Nutr

Endurance-trained men who took a ketone monoester with caffeine before 90 minutes of loaded treadmill work in 34 °C heat lasted longer in a subsequent time-to-exhaustion test than on carbohydrate, and scored better on the easier working-memory task. Harder cognitive tests, heart rate and core temperature were unchanged.

⚠What it shows: One randomised double-blind crossover trial in 17 endurance-trained young men with high VO2max; benefits appeared on some cognitive measures but not others, and the time-to-exhaustion difference was under two minutes on a short high-intensity test.RCT · J Int Soc Sports Nutr

Twelve recreationally active young men who took 1 g of Cordyceps militaris extract 30 minutes before a maximal cycling test had faster Stroop reaction times after exercise and higher resting oxygen uptake than on placebo. Peak exercise oxygen uptake, blood glucose, blood pressure and heart rate were unchanged.

⚠What it shows: Very small randomised double-blind crossover trial in 12 recreationally active young men with modest VO2max; the condition-by-time interaction was non-significant and accuracy was already at ceiling, so the finding is exploratory.RCT · PLoS One

Across 23 double-blind trials in trained cyclists, pre-exercise caffeine cut time-trial completion time with a small effect and zero heterogeneity, in both short efforts of 10 minutes or less and long ones of 20 minutes or more. Mean power output rose significantly only in the short trials.

⚠What it shows: Meta-analysis of 23 randomised, double-blind, placebo-controlled trials in healthy cyclists; effects are small (SMD −0.32 for completion time) and the subgroup difference between short and long trials was not statistically significant.Systematic review · Front Physiol

Trained men taking 90 mg per kg of fat-free mass per day of liquid HMB alongside high-intensity functional training improved a broad set of aerobic markers more than placebo, including VO2max and thresholds. Maximal isometric strength did not change with supplementation in either trained or untrained men.

⚠What it shows: One triple-blind randomised placebo-controlled trial in 53 trained and 37 untrained males over two 3-week periods with only 2 training sessions per week; no women, short duration, and the dose is far above the commonly studied 3 g/day.RCT · J Int Soc Sports Nutr

In young male taekwondo athletes, 28 days of 12 mg daily astaxanthin produced a significant group-by-time interaction for creatine kinase, a marker of muscle damage, and higher antioxidant enzyme activity. Most burnout measures were unchanged.

⚠What it shows: One 28-day randomised, double-masked, placebo-controlled trial in young male taekwondo athletes; sample size not reported in the abstract. Several of the between-group differences were exploratory, and no burnout subscale showed a significant group-by-time interaction.Study · Nutr J

Pooling eight randomised trials, movement retraining substantially reduced peak vertical ground reaction force during landing, a modifiable biomechanical risk marker for ACL injury. No included trial measured actual injury rates, so whether softer landings mean fewer torn ligaments remains unknown.

⚠What it shows: Meta-analysis of nine comparisons from eight randomised trials with only 292 participants in total; certainty of evidence was rated low with substantial heterogeneity (I2 = 63%), the subgroup difference rests on just three general-athletic comparisons, and the outcome is a surrogate marker rather than injury incidence.Review · J Funct Morphol Kinesiol

In 30 office workers with chronic neck pain, eight weeks of supervised Pilates improved pain, disability, cervical range of motion, neck muscle strength and core endurance. Adding kinesiotaping to the same programme gave no extra benefit for pain or disability, and only improved left cervical rotation.

⚠What it shows: Small randomised controlled trial with only 30 office workers completing (16 Pilates, 14 Pilates plus taping), no no-exercise control group, short 8-week follow-up and no blinding of the taping; the authors themselves call for larger studies.RCT · Pain Res Manag

Ten weeks of French contrast training beat equal-load traditional training for jump height, reactive strength, 10-metre sprint and change-of-direction performance in 25 female collegiate basketball players. Maximal strength was the exception: both groups added to their back squat one-rep max, with no advantage to either method.

⚠What it shows: Single randomised controlled trial, 25 female collegiate basketball players split into groups of 13 and 12, 10 weeks. Small groups, one sport, trained young women only: findings may not transfer to other populations or to untrained people.Study · Front Physiol

Pooling 15 randomised trials in 244 people, lower-limb blood flow restriction work produced a moderate boost in subsequent explosive performance (SMD 0.55), with the largest effects from multi-joint lifts, rest longer than 4 minutes and cuff pressures of 40-60% of arterial occlusion. It also made the effort feel significantly harder (SMD 0.82 for perceived exertion).

⚠What it shows: Meta-analysis of 15 RCTs totalling 244 participants aged 17-35, so young and mostly trained. Heterogeneity was high for the main outcome (I² = 78%) and several subgroups rested on only 1-2 studies, which the authors say should be read with extreme caution.Systematic review · Front Physiol

Eight weeks of dynamic neuromuscular stability exercises, three sessions a week, improved balance and lower-limb strength in older men: the trained group gained 3.79 points on the Berg Balance Scale, 1.77 extra repetitions on the Chair Stand Test and cut 0.87 seconds from the Timed Up and Go, versus a group doing nothing.

⚠What it shows: A small randomised controlled trial in 30 men aged 61-76 in Iran, with a no-intervention control group rather than an active comparator, so attention and general activity effects cannot be ruled out. Single site, short duration, men only.Study · J Aging Res

Pooling 22 randomised trials, a single cold-water immersion after exercise did nothing for maximal isometric strength recovery and transiently blunted jump performance immediately afterwards, while its apparent benefit for muscle-damage markers vanished once publication bias was accounted for. The one reasonably consistent effect was lower reported soreness.

⚠What it shows: Systematic review and meta-analysis of 22 RCTs in healthy people comparing one cold-water immersion bout with passive rest; certainty was moderate at best for strength and very low for soreness, with substantial heterogeneity and publication bias affecting the creatine kinase result.Meta-analysis · PeerJ

In 60 recreational CrossFit athletes aged 40 and over, both ischaemic preconditioning and cold compression lowered creatine kinase at 72 hours versus passive rest, but their effects diverged: ischaemic preconditioning showed the clearest skin-level perfusion signal while cold compression improved perceived recovery and short-term reactive strength.

⚠What it shows: Single 2-week randomised repeated-measures trial, 60 athletes aged 40+ allocated 1:1:1; only tissue perfusion was a confirmatory endpoint and the authors state the results are not evidence of endothelial function or deep muscle perfusion.Study · Sci Rep

In a Finnish twin cohort followed for a mean 17.5 years, a genetic score for longer lifespan was associated with modestly lower all-cause mortality, largely independent of lifestyle. Smoking 20 or more cigarettes a day and female sex were both stronger predictors than the genetic score.

⚠What it shows: Observational prospective cohort of 5,575 older Finnish twins (mean age 57.4) with 1,405 deaths; associations only, self-reported lifestyle factors, and findings in one national population may not generalise.Study · Geroscience

In 95 geriatric female Rottweilers, dogs that had produced litters were 3 times more likely to be in the most robust fifth of the group in old age than dogs that never bred, and those with the most litters were the most robust. The predicted "frailty cost of reproduction" did not show up at all.

⚠What it shows: Observational study in 95 geriatric female Rottweiler dogs used as a preclinical model, not in humans; robustness was defined as the lowest quintile of a 34-item frailty index, the confidence interval was wide (OR 3.30, 1.16-9.38) and selection effects, though tested, cannot be fully excluded.Study · J Gerontol A Biol Sci Med Sci

A meta-analysis of twelve stretching studies found that 4-12 week stretching programmes produced a large reduction in pulse wave velocity, a measure of arterial stiffness (SMD -1.02), while a single stretching session produced only a small, non-significant change. Repeating each stretch three or more times was linked to the biggest benefit.

⚠What it shows: Systematic review pooling just six acute and six longitudinal studies in adults, with wide confidence intervals (longitudinal SMD -1.02, 95% CI -1.79 to -0.25). The authors treat it as a basis for designing a proper efficacy trial, not proof that stretching lowers cardiovascular disease risk.Systematic review · Sports Med

Among 6313 UK Biobank adults with severe mental illness wearing accelerometers, meeting the recommended 150 minutes a week of moderate-to-vigorous activity was associated with 18% lower incident cardiovascular disease, 42% lower all-cause mortality and 49% lower cardiovascular mortality. Adding moderate light-intensity activity on top was linked to the lowest risk of all.

⚠What it shows: Prospective observational cohort, not a trial: 6313 UK Biobank participants with severe mental illness (mean age 61.05, 33.9% men), median 7 years of follow-up, with 233 deaths from any cause and 88 from cardiovascular disease. Activity was measured once by accelerometer, and reverse causation, sicker people moving less, cannot be ruled out.Study · J Am Heart Assoc

Community-dwelling adults aged 55 and over who played a home-based digital mahjong game twice weekly for six months improved slightly more on delayed recall than a no-intervention control group, with a between-group difference of 0.79 points on the Montreal Cognitive Assessment. Diastolic blood pressure also fell 3.50 mm Hg more in the game group, and brain imaging showed increased regional activity coherence in the right insula.

⚠What it shows: Single-blind randomised trial of 60 community-dwelling adults in Taipei, 30 per group, analysed per protocol, with no active control: the comparison group did nothing at all, so attention, novelty and the bundled walking incentive cannot be separated from the game itself. Cognitive gain was under one point on a screening test.RCT · JMIR Aging

A meta-analysis of 12 randomised controlled trials in 1,057 college students found Tai Chi significantly reduced depression, anxiety and stress, and also improved sleep quality. The sleep finding came with substantial variation between studies.

⚠What it shows: Meta-analysis of 12 RCTs (1,057 college students only); anxiety and sleep results showed high heterogeneity, and optimal protocols remain undefined.Meta-analysis · Front Public Health

A proteome-wide Mendelian randomisation study reports that genetically predicted branched-chain amino acid levels track with 40 circulating proteins, six of which, including PCSK9, SHBG and coagulation factor II, appear to partly mediate the link between BCAAs and ischaemic heart disease, accounting for 6.5% to 32.1% of the association. The pathways implicated involve inflammation, coagulation, lipid metabolism and cellular stress.

⚠What it shows: Genetic (Mendelian randomisation) analysis of existing large datasets, not a feeding trial: it uses genetic proxies for BCAA levels, not measured dietary BCAA or supplement intake, and mediation estimates rest on modelling assumptions. Findings are mechanistic hypotheses needing experimental confirmation.Study · J Nutr

An umbrella review of 78 meta-analyses covering 177 373 participants found app-based interventions improved blood pressure, HbA1c, fasting and post-meal glucose, bodyweight, waist circumference, physical activity, diet quality, mood and medication adherence versus controls. There was no significant effect on cardiovascular mortality, cholesterol or triglycerides, body fat, hospitalisation or smoking abstinence.

⚠What it shows: An umbrella review of existing reviews, so quality is inherited: only 1% of the 78 included reviews was rated high quality and 73% critically low, heterogeneity was very high for most outcomes, and several authors are employed by a food company.Study · Lancet Digit Health

A four-week virtual education programme about andropause, delivered over WhatsApp, significantly improved general health and sleep quality scores in retired men compared with a control group, with effects assessed again two months later.

⚠What it shows: Quasi-experimental study of just 60 retired men in Iran (30 per group), relying on self-reported questionnaires; not registered as a trial and needs replication in larger, more diverse samples.RCT · Sleep Breath