Study archive · page 43

Genetically predicted higher plasma omega-3 levels were associated with a lower risk of diabetic kidney disease (odds ratio 0.869), while omega-6 showed no significant association; a paired meta-analysis of 12 randomised trials (474 participants) found omega-3 supplementation lowered triglycerides, systolic blood pressure and a kidney injury marker, but did not improve albumin-to-creatinine ratio or estimated glomerular filtration rate.

⚠What it shows: Genetic (Mendelian randomisation) analysis plus a meta-analysis of 12 small trials totalling only 474 participants in people with diabetic kidney disease; mediation findings were nominally significant only, and hard kidney-function outcomes did not shift.Study · FASEB J

An updated meta-analysis of 35 randomised trials including 114 592 people found that new-onset atrial fibrillation risk rose only in participants at high cardiovascular risk taking more than 1500 mg/d of EPA/DHA (odds ratio 1.43, absolute risk difference 0.8%). Lower doses showed no significant increase, even in high-risk groups.

⚠What it shows: Pooled randomised trial data are strong, but this is a subgroup-defined finding across 35 heterogeneous trials in people aged 50 and over, with published and unpublished data combined; the authors call for prospective studies weighing the risk against omega-3's benefits.Study · Circ Arrhythm Electrophysiol

In a 12-week randomised, double-blind trial in 70 men aged 45-75 with male climacteric symptoms, a botanical formulation of Elaeagnus multiflora fruit and Cynanchum wilfordii extracts at 1500 mg/day was no better than placebo: symptom scores fell substantially in both arms with no significant difference between them.

⚠What it shows: One 12-week trial, 70 men randomised and 64 analysed per protocol; single formulation and dose, short follow-up, and long-term safety cannot be judged from it.Study · Medicina (Kaunas)

Pooling 17 randomised trials in non-diabetic overweight or obese adults, Lactobacillus plantarum supplementation modestly reduced body weight (-1.20 kg) and waist circumference (-1.36 cm) but did not significantly change BMI, blood lipids, fasting glucose, insulin, inflammatory markers or liver enzymes.

⚠What it shows: Meta-analysis of 17 RCTs in non-diabetic overweight or obese adults; effects are small, heterogeneity was high for BMI (I²=82.8%), and GRADE certainty limited benefits beyond weight and waist.Review · Probiotics Antimicrob Proteins

A four-arm trial in 44 men with obesity found that 12 weeks of combined aerobic-resistance training plus 3 g/day taurine produced the biggest shifts: body mass -5.50 kg, body fat -4.90%, lower asprosin and higher spexin. Training alone or taurine alone did less, and the untreated control group changed on nothing.

⚠What it shows: Four-arm randomised trial with only n=11 per group in men with a BMI of at least 30; supplement-blinded rather than fully double-blind, and the authors themselves call the results preliminary and exploratory pending larger replication.Study · Nutrients

A meta-analysis of 10 randomised trials using direct muscle imaging found that adding creatine to resistance training produces only a very small extra increase in muscle size compared with training plus placebo. Gains in upper- and lower-body muscle thickness were in the region of 0.10-0.16 cm.

⚠What it shows: Meta-analysis of 10 RCTs (44 outcomes) in healthy adults; the pooled effect was very small and its credible interval crossed zero; several authors declared ties to the creatine industry.Meta-analysis · Nutrients

Pooling placebo-controlled trials, people on GLP-1 or dual GIP/GLP-1 receptor agonists ate about 271 kcal less at a standardised ad libitum lunch, a pooled reduction of 1132 kJ with no heterogeneity between trials. The difference between semaglutide and tirzepatide was not statistically significant.

⚠What it shows: The primary pooled estimate rests on just four arms from three trials with 209 participants, and measures a single laboratory test meal rather than habitual eating; the authors note real-world dietary intake remains underreported.Review · Diabetes Metab Syndr

In an interim analysis of a 24-week trial in 47 older adults with mild cognitive impairment, a Mediterranean diet raised ketone and TCA-cycle metabolites and lowered branched-chain amino acids, while a low-fat diet did not; several of those metabolite shifts tracked with better cognitive scores.

⚠What it shows: Interim metabolomics analysis of a randomised trial in just 47 adults aged 60 and over with mild cognitive impairment, over 24 weeks. The metabolite-cognition links are associations within the trial, not proof that the diet improved cognition, and the sample is far too small to be conclusive.Study · Food Funct

Across 49 randomised trials with 1863 healthy adults, adding protein supplements to resistance training increased fat-free mass by 0.30 kg and one-rep-max strength by 2.49 kg on average. Once total protein intake passed roughly 1.62 g/kg/day, extra protein added nothing further to fat-free mass gains.

⚠What it shows: Meta-analysis of 49 RCTs in healthy adults training for at least six weeks, so the pooled effects are solid but small in absolute terms; the 1.62 g/kg/day ceiling comes from a break-point analysis of group averages, not an individualised threshold, and the senior author discloses dairy-industry grant support.Meta-analysis · Br J Sports Med

A meta-analysis of 74 randomised trials found that eating more protein adds small extra gains in lean mass and lower-body strength for healthy adults doing resistance training. Lean-mass benefits appeared at 1.2-1.59 g of protein per kg per day in adults 65 and over, and at 1.6 g/kg/day or more in younger adults.

⚠What it shows: Meta-analysis of 74 RCTs in healthy, non-obese adults; the lean-mass effect was small (SMD 0.22, moderate-level evidence), strength evidence was rated low, and some authors have food-industry ties.Meta-analysis · J Cachexia Sarcopenia Muscle

A 12-week sports vision training programme cut digital eye strain prevalence from 76.0% to 44.0% in male students and from 72.0% to 38.0% in female students, while controls given standard screen-break advice barely improved.

⚠What it shows: One RCT of 200 university students in China; symptoms were self-reported by questionnaire with no objective eye measures, so expectation effects cannot be ruled out and mechanisms remain unclear.Study · Front Neurosci

In older people with sarcopenia, resistance training and mixed training significantly improved knee-extension strength and gait speed, while whole-body vibration training improved neither. Chair-stand times did not improve with any exercise mode.

⚠What it shows: Meta-analysis of 26 studies (25 RCTs plus one non-randomised trial, 1,191 older adults with sarcopenia); heterogeneity was high in several analyses and chair-stand performance did not improve with any mode.Meta-analysis · BMC Geriatr

A network meta-analysis of 89 studies in chronic non-specific low back pain ranked Pilates most likely to reduce pain, with resistance and motor-control exercise best for physical function. Stretching and McKenzie exercise did not outperform doing nothing.

⚠What it shows: Network meta-analysis of 89 studies (5,578 patients), but the synthesised evidence was rated low quality under GRADE, with high heterogeneity and few low-risk-of-bias trials: rankings are indicative, not definitive.Meta-analysis · Br J Sports Med

In a randomised trial of 142 healthy older adults, six months of foreign-language learning kept the CA3/dentate gyrus region of the hippocampus stable while the non-language groups lost volume, whereas aerobic exercise, not language, improved memory precision on a mnemonic discrimination task. Combining both was the only way to get both effects, but the combination did not amplify either one.

⚠What it shows: One six-month randomised controlled trial in 142 healthy adults aged 63-78 with four groups, so roughly 35 people per arm; brain volume changes were not actually linked to memory changes, and the findings need replication before anyone claims a protocol for protecting the ageing hippocampus.Study · Imaging Neurosci (Camb)

In adults diagnosed with depression or anxiety, exercise produced a large improvement in depressive symptoms and a moderate improvement in anxiety symptoms across randomised trials. Aerobic, resistance and mixed programmes were all beneficial.

⚠What it shows: Meta-analysis of 26 RCTs (2,681 participants) in adults with diagnosed depression or anxiety; the anxiety analysis rested on only 11 studies, and the review could not crown a single best exercise mode.Meta-analysis · Int J Ment Health Nurs

An umbrella review pooling 133 systematic reviews and 2,724 randomised trials found exercise improves general cognition, memory and executive function across all populations and ages. Effects were generally larger for low- and moderate-intensity activity.

⚠What it shows: Umbrella review of 258,279 participants; effects were modest in size (SMD 0.24-0.42), overlap between the included reviews is inherent to the design, and review quality varied, though results held after excluding low-quality reviews.Meta-analysis · Br J Sports Med

Applying cold packs to worked muscles made no meaningful difference to soreness or strength recovery after damaging exercise: whether cooling started immediately or 24 hours later. Both cooled groups tracked the same recovery curve as controls who received no cold at all.

⚠What it shows: One RCT, n=48 healthy young adults (39 women, 9 men, mean age 22), local gel packs on one muscle only; findings may not extend to whole-body cold immersion, colder protocols or trained athletes.Study · Int J Environ Res Public Health

Adding ultrasound-guided percutaneous neuromodulation to a programme of needle electrolysis plus eccentric exercise gave no extra short-term benefit for patellar tendinopathy: pain fell substantially in both groups, with no difference between them.

⚠What it shows: Single-blinded randomised trial in only 26 people with patellar tendinopathy, followed for 21 days; small samples can miss modest real effects, and several authors have commercial ties to the clinical protocols tested.Study · Healthcare (Basel)

Depression and bipolar disorder frequently co-occur with obstructive sleep apnoea, a large pooled analysis of 23 studies finds, with older age and body-mass index significantly shaping how common the overlap is. The authors argue this supports systematic screening for one condition when the other is present.

⚠What it shows: Meta-analysis of 23 observational studies with substantial heterogeneity; prevalence data show association, not causation, and bipolar disorder numbers were far smaller than depression numbers.Review · J Clin Med

In 1001 middle-aged adults from a population-based cohort who underwent overnight polysomnography and cardiopulmonary exercise testing, a higher apnoea-hypopnoea index was associated with lower peak oxygen uptake and with all-cause mortality over a median 10.3 years (hazard ratios 1.31-1.72). Lower fitness did not statistically explain the mortality link.

⚠What it shows: Observational population cohort of 1001 adults, mean age 54, with only 73 deaths over follow-up and a single night of sleep testing: associations cannot establish cause, and the mediation analysis was underpowered.Study · ERJ Open Res

In a single-centre randomised trial of 250 middle-aged and older adults with hypertension, a structured e-coaching model delivered through a messaging app produced higher self-management and perceived-care scores after six months, plus lower systolic and diastolic blood pressure, better medication adherence and fewer hospital visits than usual care. The trial was unblinded and ran at one site.

⚠What it shows: One single-centre randomised trial of 250 middle-aged and elderly hypertension patients in China over six months; outcomes were largely questionnaire-based, participants and staff could not be blinded to a messaging-app intervention, and there is no longer-term follow-up.Study · Front Cardiovasc Med

Across 93 randomised trials, endurance, dynamic resistance and isometric resistance training all lowered resting blood pressure, with isometric training showing the largest systolic reduction at -10.9 mm Hg. Endurance training worked best in people who already had hypertension.

⚠What it shows: Meta-analysis of 93 trials (5,223 participants), but the standout isometric figure rests on only 5 training groups, so the comparison between modes is far less certain than the overall benefit.Meta-analysis · J Am Heart Assoc

A clinical review argues that exercise-induced rises in troponin and natriuretic peptides in asymptomatic athletes should be treated as mostly benign and temporary, noting that 74-96% of marathon runners exceed standard high-sensitivity troponin thresholds. The authors say sustained elevations beyond 72 hours or symptoms deserve further investigation.

⚠What it shows: Narrative clinical review, not a meta-analysis; the underlying literature it summarises is mostly small studies (most n<100) with short follow-up and inconsistent exercise protocols, and athlete-specific reference ranges do not yet exist.Review · Front Physiol

Pooling 24 observational studies and 35,916 people with type 2 diabetes, each 10% more time spent in the target glucose range was associated with lower all-cause mortality and lower risk of retinopathy, lower-limb atherosclerotic disease and peripheral and autonomic neuropathy. Associations for albuminuria and amputation did not reach significance.

⚠What it shows: Meta-analysis of 24 observational studies, so these are associations and cannot prove that raising time in range causes the lower risk; monitoring methods varied from continuous sensors to fingertip testing, and albuminuria and amputation outcomes were null.Review · J Clin Med