Strength and frailty: what the trials show
Protein, creatine, omega-3, leucine and HMB have all been tested for age-related muscle loss. The results follow one clear rule, and knowing it tells you which of them is worth anything to you.
Last scientifically reviewed: November 22, 2026Editorial review: MDS Labs. External scientific review pending.12 minute read
In brief
- One finding outranks everything else on this page: across 66 trials, physical activity was the most effective frailty intervention, and exercise plus nutrition came second. No supplement came close on its own.
- The supplements that work mostly work alongside exercise. Protein with exercise beat exercise alone; creatine did nothing without resistance training.
- The rule that explains almost all of it: the bigger the deficit, the bigger the effect. Whey protein did nothing for lean mass in older adults generally, and clearly helped those who were already sarcopenic or frail.
- HMB is the cleanest demonstration. It helped patients with diagnosed sarcopenia, and in people healthy enough to lift weights it did nothing at all.
- Omega-3 is the one exception worth knowing: at over 2.5 g a day it improved strength and walking function even in trials without exercise, while never increasing muscle mass.
What frailty and sarcopenia actually mean
Two words get used loosely, and the trials below treat them precisely.
Sarcopenia is the age-related loss of skeletal muscle, diagnosed on a combination of low muscle mass, low strength and slow walking. It is a specific clinical finding, not just being thinner than you were.
Frailty is broader: a state where small setbacks cause disproportionate harm, so a minor illness leads to a fall, a hospital stay, a loss of independence.
The measures that keep appearing are worth recognising. Handgrip strength is measured with a squeeze dynamometer and predicts a surprising amount about general health. Gait speed is how fast you walk a short measured distance. Skeletal muscle index is muscle mass adjusted for height, which is why it is reported in kg/m². Appendicular lean mass means the muscle in your arms and legs, where it matters most for getting around.
Start with the finding that outranks the rest
A network meta-analysis pulled together 66 randomized trials of interventions aimed at reducing frailty. In the core frailty analysis — 21 trials, 5,262 participants, eight different interventions compared against each other — physical activity reduced frailty with an effect size of −0.92, and the ranking put it first with 100% likelihood of being the most effective option. Physical activity combined with nutritional supplementation ranked second at 71% 1.
This is not the conclusion a supplement company would choose, and it is the honest starting point. It also makes the rest of the page more useful rather than less: once you know exercise is the engine, the question becomes which supplements make the engine run better. Several do.
Two caveats. The reviewers graded their own evidence as low to very low quality, which reflects how hard frailty trials are to run well 1. And "physical activity" in these trials usually meant a structured, supervised programme, not a resolution to walk more.
Protein, with exercise
Thirteen randomized trials in 1,057 adults aged 65 and over, all with sarcopenia or physical frailty, tested protein supplementation alone or combined with exercise 2.
What the trials show
Protein plus exercise beat exercise alone — not placebo, but exercise alone, which is the comparison that matters. Skeletal muscle index rose by 0.89 kg/m² and handgrip strength by 2.64 kg relative to exercising without the supplement 2. That is a real additive effect on top of the best available intervention.
No serious adverse events were reported across any of the trials 2.
What they do not show
Protein on its own produced only modest strength benefits and little effect on physical performance 2. The supplement is the junior partner here. Taking it instead of training is not what was tested, and is not supported.
Whey protein, and why who you are decides the answer
This analysis is the clearest illustration of the rule stated at the top, because it split its participants and got two different answers.
What the trials show
In older adults generally, whey protein had no effect on lean mass and no effect on muscle strength. It improved physical function, with gait speed in particular, but that was all 3.
In older adults who were already sarcopenic or frail, the same supplement improved lean mass (effect size 0.982), appendicular lean mass, and physical function (1.211) 3.
And adding vitamin D on top pushed it further: lean mass 0.993, muscle strength 2.005, physical function higher still 3. Muscle strength is the striking one, because whey alone had not moved strength at all.
What they do not show
The vitamin D results come from a smaller group of trials, so the large numbers carry real uncertainty. And none of this establishes that vitamin D does the same for someone whose level is already adequate — the trials recruited populations where a deficiency was plausible.
Leucine-rich protein
Leucine is one of the amino acids in protein, and the one most directly involved in signalling muscle to rebuild. Six randomized trials in 699 sarcopenic older adults tested leucine-rich supplements 4.
What the trials show
A combined improvement in strength, mass and performance of 0.939, with muscle strength on its own at 0.794 4.
What they do not show
Look at the confidence interval on that strength result: 0.104 to 1.485. The lower end is close to nothing and the upper end is very large, which means six trials is not enough to know the size of the effect, only its likely direction. The analysis also used a fixed-effects model, which assumes the trials were all estimating the same underlying number — an optimistic assumption with this few studies. Treat leucine as promising and imprecisely measured.
Creatine, and only with training
Creatine is the best-evidenced supplement in sports nutrition, and the aging research comes with a specific condition attached.
What the trials show
Combined with resistance training, creatine increased lean tissue mass and strength beyond placebo in aging adults 5.
The dosing detail is unusually actionable. A loading phase of at least 20 g a day for 5 to 7 days, followed by 5 g a day or less, improved chest press strength. Higher ongoing doses above 5 g a day improved leg press strength 5.
What they do not show
When trials using a loading phase were excluded, creatine showed no greater effect than placebo on chest or leg press strength 5. The loading phase appears to be doing meaningful work rather than being a marketing ritual.
And the whole result is conditional on resistance training. This is not a supplement that builds muscle while you wait.
Omega-3, the one that works without exercise
A network meta-analysis of 16 randomized trials in people with sarcopenia or at high risk of it compared different omega-3 doses against each other 6.
What the trials show
High dose, meaning more than 2.5 g a day, produced the largest improvements in both upper-body muscle strength (1.68) and lower-body physical function (0.73). Crucially, the effects held up in a separate analysis of only the trials that had no resistance training programme attached 6. That makes omega-3 the one item here with evidence of working on its own.
What they do not show
No omega-3 regimen significantly increased skeletal muscle mass 6. People got stronger and moved better without getting measurably more muscle, which is a useful reminder that strength and size are not the same thing.
The strength result also has a confidence interval of 0.03 to 3.33. The lower bound is barely distinguishable from zero. The function result (0.16 to 1.30) is the more trustworthy of the two.
HMB, and the lesson hidden in a contradiction
HMB is a compound your body makes from leucine, sold specifically for preserving muscle. Two meta-analyses appear to disagree about it, and the disagreement is the most instructive thing on this page.
What the trials show
In patients with clinically diagnosed sarcopenia, five trials found HMB improved skeletal muscle mass index (0.32) and handgrip strength (0.65) 7.
What they do not show
In adults aged 50 and over who were able to do structured resistance training, 13 trials in 561 people found HMB added nothing: fat mass 0.24, muscle mass 0.05, muscle strength 0.04 — all statistically indistinguishable from placebo. The authors concluded it cannot be recommended as a routine adjunct to resistance training for people able to exercise 8.
Even the positive finding is thin. Both of its results sat right at the edge of significance (P = 0.048 and P = 0.034), the muscle mass interval touched zero exactly, and gait speed did not improve at all 7.
These two results are not actually in conflict. They tested different people. HMB appears to do something when there is a real deficit to correct, and nothing when there is not.
The doses and conditions that were actually studied
| What | What improved | The condition attached |
|---|---|---|
| Structured physical activity | Frailty itself — ranked first of 8 interventions | Supervised programmes, low-quality evidence |
| Protein | Muscle index +0.89 kg/m², grip +2.64 kg | Only when added to exercise |
| Whey protein | Lean mass and function | Only in sarcopenic or frail people |
| Whey + vitamin D | Lean mass, strength, function — all larger | Fewer trials, wider uncertainty |
| Leucine-rich protein | Strength in sarcopenia | 6 trials, very imprecise |
| Creatine, 20 g/day load then ≤5 g/day | Lean mass and strength | Only with resistance training; loading needed |
| Omega-3, >2.5 g/day | Strength and walking function, not mass | Works without exercise |
| HMB | Muscle mass and grip in diagnosed sarcopenia | Nothing in people who can train |
How to read an effect size here
Most results above are reported as SMD, standardized mean difference, which lets trials using different measurements be pooled. Roughly: 0.2 is small, 0.5 moderate, 0.8 large.
Two habits make these readable. Look at the interval, not just the number. An effect of 1.68 sounds enormous, but with an interval running from 0.03 to 3.33 the trials have really only established that the effect is probably positive. Compare that with protein's +2.64 kg of grip strength, which is a concrete quantity in real units.
Check what the comparison group was doing. This matters more here than anywhere else in the Lab. "Better than placebo" and "better than exercise alone" are very different claims, and the protein result is the stronger kind.
What this adds up to
The evidence here is more coherent than it first looks, and it points somewhere specific.
Exercise, particularly resistance training, is the intervention. It ranked first across 66 trials, and it is the precondition for creatine and the partner for protein. Anyone treating supplements as a substitute has the structure backwards.
On top of that, the additions with real support are protein, which improved muscle and grip strength beyond exercise alone, and creatine with a proper loading phase. For someone who is already sarcopenic or frail, whey with vitamin D has the largest effects on the page. For someone who genuinely cannot train, high-dose omega-3 is the one option with evidence of working unaided, for strength and walking rather than muscle size.
And the rule worth carrying away is the one the HMB and whey results both demonstrate: these supplements work in proportion to what is missing. That is good news if you are frail and the reason to be skeptical of the marketing if you are not. A product promising to protect the muscle of a healthy, active 55-year-old is making a claim that the trials in healthy, active people did not support.
Keep reading
Inflammation runs alongside muscle loss in aging, and is covered in Inflammation: what the trials show. For why dose and formulation decide whether a supplement does anything at all, see Bioavailability and half-life, and for reading trial evidence generally, Understanding evidence.
Two of our research articles follow this thread further: Bazi Bushen: A Randomized Human Trial on Aging, and A Three-Compound Blend Lowered Immune Biological Age.
Unexplained weakness, weight loss or a fall has causes worth investigating, and several are treatable. That is a conversation with a doctor first. Creatine loading doses and high-dose omega-3 both warrant a check with a pharmacist if you have kidney disease or take a blood thinner.
