Research library
It also includes studies that show something other than expected. A library where everything confirms its own story isn't a library, it's a sales pitch.
State of the library
By pillar
Two ways to use this library
“I mainly want to know what to do.”
Practice-focused summaries: the conclusion, what it means for your training or nutrition, little methodology.
“I want to look at the underlying studies.”
For anyone who wants the methodology, limitations, evidence rating and direct source link — via each study's detail page.
The library is organised by the five pillars and holds 170 selected publications. This collection keeps growing: new research is added once it genuinely adds to or contradicts existing knowledge.
170 studies · 25 of 170 shown
2026
Resistance training improved strength, hypertrophy, power, endurance, contraction velocity, gait speed, balance, and physical function. For strength: heavier loads (≥80% 1RM), full range of motion, 2-3 sets, start of session, ≥2x/week. For hypertrophy: higher volume (≥10 sets/week) and eccentric overload. Training to failure, equipment type, periodization did not consistently affect outcomes.
2026
Combining strength and endurance gave the same fitness gains as endurance alone, plus more strength. Compared with strength training alone, fitness improved more and strength and muscle growth were similar.
2026
Creatine added on average 0.37 kg of lean mass and 7.5 kg of leg press strength versus placebo, mainly at 5 g a day or more with resistance training. Bone density was unchanged.
2026
Possibly somewhat fewer falls (rate ratio 0.76) than with usual care, but no added benefit over falls advice or education, and uncertain compared with exercise.
2026
Partly: less muscle soreness up to 96 h and better muscular power after 24 h, but no better muscular strength, and with chronic use smaller gains in strength and muscle mass.
2026
A little: strength gains are somewhat larger with creatine (g = 0.31), but no demonstrable difference was found for muscle mass or everyday physical function.
2026
Possibly: on average a positive effect on bone measures, but with very low certainty and large differences between programmes. Whether it reduces wrist fractures was not studied.
2026
A little: small improvements in, among others, distance covered (g = 0.17) and workload (g = 0.25), but no effect on muscular endurance, maximum power output or VO2max.
2026
Adding sprint intervals to strength training gave the same gains in strength and jump performance as strength training alone. Maximum oxygen uptake rose more.
2026
Volume and frequency have unique dose-response relationships with hypertrophy and strength. Volume showed diminishing returns for both; strength had more pronounced flattening. Frequency affected strength (with diminishing returns) but not consistently hypertrophy. Analysis of 67 studies, 2058 participants (mean age 25 years, 79% men).
2026
People who ate less and also exercised lost on average 0.87 kg less fat-free mass than those who only ate less. Exercise prevented nearly half (45.7%) of that loss.
2026
Eccentric training, which emphasises lowering the weight, gave slightly greater strength gains than traditional training (SMD 0.27). Function, power and muscle growth did not differ.
2026
Yes, but the association is small: women with a higher protein intake had slightly higher handgrip strength on average (SMD 0.15, 95% CI 0.10 to 0.20).
2026
All training conditions produced more muscle mass and strength than no training. Training extra hard in one specific cycle phase gave no advantage over evenly spread training.
2026
All groups lost about the same amount of weight. Resistance training gave the largest fat loss and was the only type where fat-free mass (everything except fat) increased, in men and women.
2026
Neither whey nor soy protein significantly changed lean body mass. Whey protein did improve bench press (mean difference 8.87) and squat (mean difference 9.60).
2026
No diet type is clearly superior; what counts is the deficit, with resistance training and sufficient protein (for the general population 1.2-1.6 g per kg body weight per day) to preserve lean mass.
2026
Whey protein alongside strength training gave the largest gain in lean body mass, just over a kilogram more than placebo. The gain appeared to level off at 30 to 35 grams per day.
2025
After 12 weeks, maximal and explosive strength increased in all groups, even when interval training and strength training targeted the same muscles. There was no clear difference between groups.
2025
Barely: faster and slower lifting produced comparable muscle growth, with a trivial difference of 0.09 (95% credible interval −0.04 to 0.22).
2025
No statistically significant hypertrophy difference between eccentric and concentric actions (ES 0.285, 95% CI -0.131 to 0.701, p=0.179). Subgroup: eccentric better for upper limb muscles (p=0.018), interventions ≤8 weeks (p=0.046), muscle thickness measures (p=0.035), isokinetic contractions (p=0.025). 26 studies, 682 participants.
2025
Static stretching made adults more flexible. Gains peaked at 4 minutes per session or 10 minutes per week; stretching more gave no extra gain.
2025
A low training volume gave the largest gains in walking and get-up tests, lean mass and muscle growth. Greater leg strength required a moderate or high volume.
2025
In studies of at least 10 weeks, people doing aerobic or combined training lost just over 1 kg more fat mass than with strength training alone. Body fat percentage did not differ.
2025
With supersets, participants completed the same repetitions and training volume in less time. Strength and muscle growth (hypertrophy) increased equally over time. Sessions felt harder.
25 of 170 shown
A study isn't good because it's large, and it isn't wrong because it's small. What counts is whether the design can answer the question.
We look at study design, because a randomised trial says something different from an observation. At the control group, because progress without comparison proves nothing. At the size and spread of the effect, because a statistically significant difference can be negligible in practice. And at the participants, because a finding in trained twenty-somethings doesn't automatically apply to a fifty-year-old just starting out.
Systematic reviews and meta-analyses of randomised trials therefore carry more weight than single studies. Position stands from scientific associations carry substantial weight because they've been appraised by multiple researchers. We don't include animal research where the translation to humans is unclear. Research funded by a party with an interest in the outcome is only included with that noted.
What we don't do is select only the sources that suit us. Where the literature is divided on a point, that's stated.
High confidence · 55
Multiple consistent meta-analyses or a position stand. Sufficient basis for a fixed choice in the coaching.
Reasonable confidence · 93
A solid systematic review with caveats, or consistent indications with a limited scope.
Interpret with caution · 21
Limited, mixed or hard-to-translate evidence. Useful as direction, not as a rule.
Not usable as a basis · 1
Insufficient quality or insufficient applicability to our target group. Included for illustration or refutation.
This library isn't an appendix. Its contents determine how programmes are built and where the limits sit on what we're willing to promise. Where the evidence is strong, the approach is definite. Where it's thin, we leave room for the idea that it could be different.
This information is general education about scientific research. It is not personal medical advice and does not replace diagnosis or treatment by a doctor or other healthcare professional. If you have symptoms, a medical condition or take medication, consult your GP or treating professional before changing your training or nutrition.
