Did people actually get stronger, or just bigger?

What this archive is for. Trials of the compound come first, ranked by how many people were actually given it. Studies that only observed people are below them and labelled.

1. 01MelittinPeptideModerate. At least one finished study in people, and at least 100 people who took it. 5 completed trials, 658 people given it.4 reported

   538 people given it1 trial

   • This tested whole bee venom, not melittin. Five hundred and thirty eight people with knee arthritis got weekly skin injections of venom, or of a control substance, for twelve weeks. Pain and physical function improved more on the venom. The gains were modest in size. Reactions where the needle went in were uncommon. Melittin is one ingredient of bee venom, and nothing here tells you what melittin alone would do. The paper does not say who paid.

     The numbers Whole purified venom, not melittin. 538 patients with knee osteoarthritis randomised 1:2 to histamine control or 100 mcg venom in 15 dermal injections at acupuncture points weekly for 12 weeks. WOMAC pain improved 1.1 points more than control (95% CI 0.3 to 2.0, p = 0.001) and physical function 3.1 points (p = 0.0046). Injection site reactions under 5 percent. Described as a phase 3; funding not stated in the abstract.

     Journal of Alternative and Complementary Medicine, 2019· 538 people

2. 02ElamipretidePeptideModerate. At least one finished study in people, and at least 100 people who took it. 5 completed trials, 487 people given it.6 reported

   10 people given it1 trial

   • A handful of patients kept improving for years, and there was nobody to compare them against. Ten people with Barth syndrome stayed on elamipretide in an extension where everybody got the drug. Their six minute walking distance rose by 96.1 metres over 168 weeks, a bit over three years. Reactions where they injected were the most common side effect. With no untreated group, growing older and getting practised at the walking test cannot be separated from the drug. Ten people is also too few to catch rare harms.

     The numbers Ten Barth syndrome patients in the open-label extension: 6-minute walk distance improved cumulatively by 96.1 m at week 168 with injection-site reactions the most common adverse event.

     Genetics in Medicine, 2024· 10 people

3. 03MGFPeptidePreliminary. Studied in people, but no published study has had people take it. 1 human study where nobody was given it.6 reported

   nobody was given it1 watch-only study1 found nothing

   • found nothingThis is the study sold as human evidence for MGF, and nobody in it was given MGF. 8 younger men and 7 men in their seventies and eighties did heavy one-leg knee extensions, then had a small piece of thigh muscle taken. The younger men's muscle switched on more of the MGF gene message. The older men's did not. This measures what the body does after hard exercise. It says nothing about injecting a lab-made copy.

     The numbers 8 men aged 25 to 36 and 7 aged 70 to 82 did 10 sets of 6 single-leg knee extensions at 80% of one-repetition maximum. MGF mRNA rose in the young group and not in the elderly group, IGF-IEa did not change in either, and resting MGF was roughly 100-fold lower than IGF-IEa. This is a study of the body's own transcription, not of an injected peptide.

     Journal of Physiology, 2003· 8 peopleobserved only, nobody was given it

4. 04MOTS-c with SS-31BlendPreliminary. Studied in people, but no published study has had people take it. 1 human study where nobody was given it.6 reported

   nobody was given it1 watch-only study

   • The main study of elamipretide failed. This is a later second look at its data, hunting for a group that did better. Across all the adults with an inherited mitochondria muscle disease, the drug showed no clear benefit. Sorted by genetic type afterwards, one type walked about 25 metres further in six minutes than people on the dummy. A smaller type walked about 37 metres further. Sorting after the fact cannot prove anything. It was meant to shape the next study.

     The numbers As previously published, MMPOWER-3 did not demonstrate a significant benefit of elamipretide in a genotypically diverse adult population with primary mitochondrial myopathy. In post hoc analysis the mitochondrial DNA replisome cohort improved 25.2 plus or minus 8.7 metres on the six-minute walk test against 2.0 plus or minus 8.6 metres on placebo at p = 0.06, and the subset with chronic progressive external ophthalmoplegia improved 37.3 plus or minus 9.5 metres against a 8.0 metre decline at p = 0.0024. These analyses were designed to inform a follow-up phase 3 trial.

     Orphanet Journal of Rare Diseases, 2024· number of people not givenobserved only, nobody was given it

5. 05Mitochondrial+ blendBlendPreliminary. Studied in people, but no published study has had people take it. 1 human study where nobody was given it.6 reported

   nobody was given it1 watch-only study

   • Elamipretide's main study did not work, and these numbers come from digging through the results afterwards. Across all the adults with an inherited mitochondria muscle disease, there was no clear benefit. Sorted by genetic type after the fact, one group walked about 25 metres further in six minutes than the dummy group. Another walked about 37 metres further. Hunting for winning groups once the data is already in hand is how false findings get made. None of this is proof until a new study tests it.

     The numbers MMPOWER-3 as published did not demonstrate a significant benefit of elamipretide in a genotypically diverse adult population with primary mitochondrial myopathy. Post hoc, the mitochondrial DNA replisome cohort improved 25.2 metres on the six-minute walk against 2.0 metres on placebo at p = 0.06, and the chronic progressive external ophthalmoplegia subset improved 37.3 metres against a 8.0 metre decline at p = 0.0024.

     Orphanet Journal of Rare Diseases, 2024· number of people not givenobserved only, nobody was given it

Vitamins, minerals, amino acids and botanicals sold over the counter, held to the same standard.

1. 01CarnosineSupplementModerate. At least one finished study in people, and at least 100 people who took it. 2 completed trials, 299 people given it.4 reported

   299 people given it1 trial1 found nothing

   • found nothingCarnosine did nothing for physical performance overall, and the effects reported were found by slicing the group up afterwards. 299 people took 2 grams a day or a dummy and were tested on grip strength, calf raises, a two minute step test and walking speed over about twelve weeks. Nothing separated the groups on any main measure. Two scattered results turned up in subgroups: more calf raises in the under 40s, and more steps in men over 40. The authors say any benefit looked limited to men.

     The numbers 299 participants randomised to placebo or 2 g a day of carnosine, with hand grip strength, bilateral calf raise, two-minute step test and gait speed measured at baseline and at roughly 6 and 12 weeks. Results were subgroup-only: a significant increase in calf raises at visit 3 in participants under 40 (P equal to 0.018), and a significant increase in step count at visit 4 in males over 40 (P equal to 0.010). No significant differences were observed on any other physical performance measure, and the authors state the benefits appeared to be limited to males.

     Journal of the International Society of Sports Nutrition, 2026· 299 people

2. 02EcdysteroneSupplementModerate. At least one finished study in people, and at least 100 people who took it. 3 completed trials, 233 people given it.5 reported

   233 people given it2 trials1 found nothing

   • found nothingThe drug version of this molecule was tested in 233 older adults losing muscle with age. It missed its main target. Across everyone who took part, walking speed over a 400 metre course barely changed compared with a dummy pill. A smaller group who stuck to the plan walked a little faster. Counting only those people is not the fair test. Full detail is on the BIO101 page.

     The numbers The pharmaceutical-grade version of this molecule in 233 older adults with sarcopenia. The primary endpoint, gait speed on a 400-metre walk, improved 0.07 m/s against placebo in the full analysis set and was not significant, and improved 0.09 m/s in the per-protocol population at p = 0.008. Detail on the BIO101 page.

     Journal of Cachexia, Sarcopenia and Muscle, 2025· 233 people

   • In a study in people, those taking ecdysterone gained more muscle than the comparison group and lifted more on the bench press. Blood tests for liver and kidney damage did not rise. The researchers checked everyone for banned drugs and tested the supplement for hidden steroids first. They said it should go on the banned list in sport. That is how strong they judged the effect. In a dish, the same growth showed up in muscle cells.

     The numbers A controlled clinical trial in humans. Significantly higher increases in muscle mass in participants dosed with ecdysterone, and significantly more pronounced increases in one-repetition bench press performance. The same hypertrophic effect appeared in C2C12 myotubes in vitro. No increase in liver or kidney toxicity biomarkers. Participants were screened for prohibited performance-enhancing substances and the supplement was tested for anabolic steroid contamination before administration. The authors conclude their results strongly suggest including ecdysterone on the prohibited list in class S1.2, other anabolic agents.

     Archives of Toxicology, 2019· number of people not given

3. 03CoQ10SupplementModerate. At least one finished study in people, and at least 100 people who took it. 2 completed trials, 162 people given it.4 reported

   120 people given it1 trial1 found nothing

   • found nothingEven in people whose statin muscle pain was confirmed by taking them on and off the statin, CoQ10 did not help. 41 of those patients then took 600 mg a day of ubiquinol with their statin, or a dummy, for 8 weeks. Pain scores rose on the statin either way. Strength and fitness did not change. Blood CoQ10 did rise, so the pills were absorbed. Absorption is not a result.

     The numbers Statin myalgia first confirmed in 120 patients by an 8-week simvastatin versus placebo crossover. 41 who hurt on simvastatin and not on placebo were then randomised to simvastatin plus ubiquinol 600 mg/day or placebo for 8 weeks. Pain severity and interference rose on simvastatin regardless of CoQ10 (p = 0.53 and 0.56). Muscle strength and VO2max did not change.

     Atherosclerosis, 2015· 120 people

4. 04Nicotinamide RibosideSupplementLimited. People have taken it in a published study, but fewer than 100 of them. 4 completed trials, 90 people given it.5 reported

   90 people given it1 trial

   • For the first time, this supplement changed something a person could actually do. 90 patients with poor circulation in their legs took it for six months. They could walk about 17.6 metres further in six minutes than the dummy group, which is around 20 extra paces. That is modest, and it is real. It is also the only result this compound has that is about function rather than chemistry.

     The numbers 90 patients: 6 months of NR improved 6-minute walk distance by 17.6 m versus placebo, a first functional outcome.

     Nature Communications, 2024· 90 people

5. 05NMNSupplementLimited. People have taken it in a published study, but fewer than 100 of them. 4 completed trials, 94 people given it.6 reported

   80 people given it2 trials2 found nothing

   • found nothingNMN raised NAD in the blood and added a little to how far people could walk, and did nothing for blood sugar handling. Eighty healthy middle aged adults took one of three daily doses or a dummy for two months, and nobody involved knew which. Walking distance in six minutes improved modestly. Insulin resistance did not budge. Most people took it without trouble. Two months is short, walking distance is easy to improve, and nobody was followed long enough to see whether anything lasted.

     The numbers 60-day RCT in 80 adults: 300 to 900 mg/day raised blood NAD and modestly improved six-minute walk distance; well tolerated; HOMA-IR unchanged.

     GeroScience, 2023· 80 people

   • found nothingA high dose NMN tablet raised NAD in the blood and moved weight and cholesterol, but not muscle. 30 overweight or obese adults took MIB-626, 1000 mg twice a day, or a dummy pill, for 28 days. Weight fell 1.9 kg. The lower blood pressure number and LDL cholesterol fell too. Strength, tiredness, fitness and stair climbing power did not change. Neither did the body's response to insulin. The maker paid for it.

     The numbers 30 overweight or obese adults aged 45 and over randomised 2:1 to MIB-626 (microcrystalline NMN) 1000 mg twice daily or placebo for 28 days. Weight -1.9 kg, diastolic BP -7.01 mmHg, LDL -18.73 mg/dL versus placebo. Muscle strength, fatigability, aerobic capacity and stair-climbing power did not differ. Insulin sensitivity unchanged. Funder: Metro International Biotech.

     Journal of Clinical Endocrinology and Metabolism, 2023· number of people not given

6. 06L-leucineSupplementLimited. People have taken it in a published study, but fewer than 100 of them. 1 completed trial, 30 people given it.3 reported

   30 people given it1 trial1 found nothing

   • found nothingThree months of leucine did nothing at all for muscle in healthy older men. 30 men averaging 71 took 2.5 g of leucine with each main meal, 7.5 g a day, or a dummy, decided by chance. Scans of muscle size showed no change in either group. Strength on a one-lift maximum did not change either. Insulin sensitivity, long-run blood sugar and blood fats were all unmoved. These men were healthy to begin with, not frail and not already losing muscle.

     The numbers 30 healthy men, mean age 71, randomised to 2.5 g leucine or placebo with each main meal, 7.5 g daily for three months. No change in skeletal muscle mass measured by computed tomography and DXA, and no change in one-repetition-maximum strength, in either group. No improvement in whole-body insulin sensitivity, HbA1c or plasma lipids.

     American Journal of Clinical Nutrition, 2009· 30 people

7. 07Urolithin ASupplementLimited. People have taken it in a published study, but fewer than 100 of them. 2 completed trials, none stating how many people.4 reported

   0 people given it1 trial, 1 watch-only1 found nothing

   • found nothingFour months of it improved muscle endurance and blood chemistry, and left the headline measures untouched. Older adults took the supplement or a dummy. How far they could walk in six minutes did not differ, and neither did the peak energy their muscles could produce. What improved was how long a muscle could keep going, plus several markers. A genuine result, and a narrow one.

     The numbers 4 months of urolithin A improved muscle endurance and plasma biomarkers, though 6-minute walk and peak ATP did not differ from placebo.

     JAMA Network Open, 2022· number of people not given

   • Strength improved, and the trial still missed the thing it was built to show. Muscle strength in middle-aged adults rose about 12 percent, and fitness and walking distance improved as well. But peak power, the one measure chosen before the trial began, did not. When a study misses the measure it named first and hits others instead, those others carry less weight.

     The numbers About 12% gain in muscle strength and improved VO2 peak and 6-minute walk, though the primary peak-power endpoint was not met.

     Cell Reports Medicine, 2022· number of people not givenobserved only, nobody was given it

   What people using it report

   • The most detailed negative account describes four months of daily use of the branded powder with no change in strength, endurance or power output, and frames the spend of several hundred dollars as wasted. Another describes overwhelming fatigue while taking it. A third describes a clean lift in energy.
8. 08QuercetinSupplementLimited. People have taken it in a published study, but fewer than 100 of them. 3 completed trials, 60 people given it.2 reported

   0 people given it1 trial, 1 watch-only

   • Everyone here knew they were getting the drugs, and they walked better afterwards. 14 patients with lung scarring took part. Distance covered in six minutes, walking speed and time to rise from a chair all improved. There was nobody to compare them against. People who join a trial and receive close attention often improve anyway, which is precisely why the same team later added a dummy group.

     The numbers 14 patients: D+Q was feasible and improved 6-minute walk, gait speed and chair-stand time; no control group.

     EBioMedicine, 2019· 14 peopleobserved only, nobody was given it

   • When the same team repeated their pilot with a dummy group added, the benefit vanished. Twelve people took part. Frailty, lung function and physical function came out much the same whichever they were given. Side effects were far more common on the drug, 65 reports against 22. Twelve people is tiny. The difference between the two studies is still the thing to notice.

     The numbers The same team's open-label pilot repeated with a placebo arm, n=12: frailty, pulmonary and physical function did not appear to differ meaningfully between groups, and the drug arm reported 65 non-serious adverse events against 22 on placebo.

     EBioMedicine, 2023· number of people not given

9. 09GlyNACSupplementLimited. People have taken it in a published study, but fewer than 100 of them. 2 completed trials, none stating how many people.3 reported

   0 people given it1 trial

   • Older adults improved on several fronts, in a study with nobody to compare them against. Over 24 weeks the supplement corrected a shortage of the body's main antioxidant, and strength, walking speed and thinking all improved. The benefits then faded once people stopped, which at least points back at the supplement. Everybody knew they were taking it, and there was no dummy group.

     The numbers 24 weeks of GlyNAC in older adults corrected glutathione deficiency and improved strength, gait speed and cognition; benefits faded after stopping.

     Clinical and Translational Medicine, 2021· number of people not given

10. 10MIB-626SupplementModerate. At least one finished study in people, and at least 100 people who took it. 4 completed trials, 107 people given it.1 reported

    0 people given it1 trial1 found nothing

    • found nothingThe tablet moved some body measurements and left every measure of physical function untouched. 30 overweight or obese adults aged 45 and over took MIB-626 or a dummy tablet for four weeks. Weight fell by about two kilograms. Blood pressure and cholesterol both came down. Strength, stamina, how quickly muscles tired and stair climbing power did not change. Insulin sensitivity, liver fat and belly fat did not change either. The manufacturer helped fund it. This study looked at many things at once, so treat any single result as a hint.

      The numbers Randomised 2:1, 30 overweight or obese adults aged 45 and over, MIB-626 two 500 mg tablets twice daily for 28 days. Body weight fell 1.9 kg (95 percent CI -3.3 to -0.5, P=.008), diastolic blood pressure fell 7.01 mmHg (-13.44 to -0.59, P=.034), total cholesterol fell 26.89 mg/dL (P=.004) and LDL fell 18.73 mg/dL (P=.007). Muscle strength, muscle fatigability, aerobic capacity and stair-climbing power did not change. Insulin sensitivity, hepatic fat and intra-abdominal fat did not change in either group. Metro International Biotech, the manufacturer, is a listed funder. This is a physiologic study with multiple outcomes rather than a trial powered on one of them.

      Journal of Clinical Endocrinology and Metabolism, 2023· number of people not given

11. 11Urolithin BSupplementLimited. People have taken it in a published study, but fewer than 100 of them. 1 completed trial, none stating how many people.

    0 people given it1 trial1 found nothing

    • found nothingThis is urolithin A, not urolithin B, and it sits here because urolithin B has nothing like it. Older adults took urolithin A or a dummy. Muscle endurance improved and some blood markers moved. The six minute walk test did not improve, and neither did the peak energy output measured inside muscle. So even the better studied sibling comes out mixed. Urolithin B has never been tested this way, or in any other way, in a person.

      The numbers Included as the contrast, and it is not urolithin B. This is a randomised trial of urolithin A in older adults, in which muscle endurance and plasma biomarkers improved while the six-minute walk and peak ATP did not differ from placebo. Urolithin B has no trial of this kind, or of any kind.

      JAMA Network Open, 2022· number of people not given

12. 12EpicatechinSupplementLimited. People have taken it in a published study, but fewer than 100 of them. 4 completed trials, 44 people given it.1 reported

    nobody was given it1 watch-only study

    • This is the study most supplement marketing traces back to, and it cannot carry that weight. Alongside mouse work, a first look in people gave epicatechin for seven days and reported stronger hand grip and a shift in two muscle-signalling proteins. Nobody went without the compound, so there is nothing to compare against. Seven days is not long enough for muscle to change. Treat it as a first hint, not a result.

      The numbers Mouse data plus an initial proof-of-concept in humans: 7 days of (-)-epicatechin increased hand grip strength and the plasma follistatin to myostatin ratio. Small, uncontrolled and short; this is the study most supplement marketing traces back to.

      Journal of Nutritional Biochemistry, 2014· number of people not givenobserved only, nobody was given it

Approved drugs, biologics and hormones that have been measured for this. They are here because leaving them out would misrepresent the evidence, and they are below the peptides because that is not what most people come here to read.

1. 01ErythropoietinBiologicStrong. At least two finished studies in people, and at least 1,000 people who took it. 3 completed trials, 4,641 people given it.6 reported

   603 people given it1 trial

   • Aiming for a normal blood count did not help, and more of those patients ended up on dialysis. 603 people with failing kidneys were treated towards a normal or a lower blood count target and followed for three years. Heart events came out about the same, 58 against 47, a gap that could be chance. More people in the normal target group needed dialysis, 127 against 111. General health and physical function scores did improve. Roche, which sells the drug, ran it.

     The numbers CREATE: 603 people with estimated GFR 15 to 35 randomised to a normal (13.0 to 15.0 g/dL) or subnormal (10.5 to 11.5 g/dL) haemoglobin target using epoetin beta, 3 years. First cardiovascular events 58 versus 47, hazard ratio 0.78, 95 percent confidence interval 0.53 to 1.14, p = 0.20. More people in the normalisation group needed dialysis, 127 versus 111, p = 0.03. General health and physical function scores improved. Sponsor Hoffmann-La Roche.

     New England Journal of Medicine, 2006· 603 people

2. 02BIO101Small moleculeModerate. At least one finished study in people, and at least 100 people who took it. 1 completed trial, 233 people given it.3 reported

   233 people given it1 trial1 found nothing

   • found nothingIt missed the target it was built to hit. 233 older adults losing muscle with age, average age 75, took BIO101 at one of two doses or a dummy pill for six months. The measure was walking speed over 400 metres. Counting everyone, the drug made almost no difference. Covid wiped out more than half the final in-person check-ups, which made any effect harder to spot. Many authors work for or advise the company that makes it.

     The numbers SARA-INT. 233 randomised, mean age 75.5, 54.3% female, three arms at 175 mg twice daily, 350 mg twice daily or placebo, planned six months and up to nine in 50 subjects. Eligibility required meeting FNIH sarcopenia criteria and a Short Physical Performance Battery score of 8 or less out of 12. Primary endpoint was change in gait speed on the 400-metre walk test. At 350 mg twice daily, gait speed improved 0.07 m/s against placebo in the full analysis set, not significant, and 0.09 m/s in the per-protocol population, p = 0.008. COVID-19 cost 55% of on-site end-of-treatment efficacy assessments, reducing power. Predefined higher-risk subpopulations showed effects of 0.047 to 0.066 m/s with a trend toward dose response. Safety was good, with related treatment-emergent adverse events in 16.0%, 13.3% and 13.5% of the placebo, 175 mg and 350 mg groups. Many authors are employees, former employees, advisory board members or consultants of the sponsor, Biophytis.

     Journal of Cachexia, Sarcopenia and Muscle, 2025· 233 people

3. 03LandogrozumabBiologicModerate. At least one finished study in people, and at least 100 people who took it. 3 completed trials, 726 people given it.

   201 people given it1 trial1 found nothing

   • found nothingA myostatin antibody added muscle to older people who fall, and barely moved how they function. Two hundred older adults were split by chance between the drug and a dummy injection. After six months the drug group had gained about four hundred grams of arm and leg muscle, while the dummy group had lost a little. Climbing twelve stairs got slightly faster. Most other movement tests did not separate. Injection site reactions were far more common on the drug. Eli Lilly paid for it.

     The numbers 201 older adults randomised, 99 to placebo and 102 to the antibody, from 365 screened. At 24 weeks appendicular lean body mass changed by minus 0.123 kg on placebo and plus 0.303 kg on drug, a difference of 0.43 kg (95% CI 0.192 to 0.660, p < 0.0001). On function: twelve-step stair climb minus 1.28 s (p = 0.011), four-step stair climb minus 0.46 s (p = 0.093), chair rise with arms minus 4.15 s (p = 0.054), fast gait speed plus 0.05 m/s (p = 0.088), tested against a prespecified two-sided alpha of 0.1 for performance tests. No effect on other performance measures. Injection site reactions in 30% on drug against 9% on placebo (p < 0.0001). Funded by Eli Lilly.

     Lancet Diabetes and Endocrinology, 2015· 201 people

4. 04DHEAHormoneModerate. At least one finished study in people, and at least 100 people who took it. 3 completed trials, 267 people given it.5 reported

   144 people given it1 trial1 found nothing

   • found nothingThe biggest study of DHEA against a dummy pill as an anti-ageing supplement found nothing that mattered. 144 older men and women with low levels took it for two years. The hormone went up and little else changed. Body shape, strength, fitness, blood sugar control and quality of life did not get better. Bone density went up slightly at one spot in men and one in women. These were healthy volunteers with low levels, not patients.

     The numbers 144 people, 87 men and 57 women aged 60 or over with low DHEA-S, randomised double-blind to 2 years of oral DHEA, testosterone (men only) or placebo. DHEA raised serum DHEA-S by a median 3.4 mcg/mL in men and 3.8 in women. No significant effect on body composition, peak oxygen uptake, muscle strength, insulin sensitivity or quality of life. Small gains in bone density at the femoral neck in men and the ultradistal radius in women. No major adverse effects.

     The New England Journal of Medicine, 2006· 144 people

5. 05Sodium phenylbutyrateSmall moleculeModerate. At least one finished study in people, and at least 100 people who took it. 2 completed trials, 148 people given it.4 reported

   137 people given it1 trial

   • In ALS, a pill combining sodium phenylbutyrate with a bile acid slowed the loss of daily function a little over six months, compared with a dummy pill. 137 people took part. The other measures of strength, breathing and survival did not differ. This small trial led to an approval that was pulled after a larger trial failed.

     The numbers 137 people with definite ALS randomised 2:1 double-blind to sodium phenylbutyrate 3 g plus taurursodiol 1 g, or placebo, for 24 weeks (CENTAUR). ALSFRS-R declined 1.24 versus 1.66 points per month (difference 0.42, 95% CI 0.03 to 0.81, p=0.03). The secondary outcomes (muscle strength, neurofilament, slow vital capacity, survival composites) did not separate from placebo. Adverse events were mainly gastrointestinal.

     The New England Journal of Medicine, 2020· 137 people

6. 06StamulumabBiologicModerate. At least one finished study in people, and at least 100 people who took it. 2 completed trials, 122 people given it.

   122 people given it2 trials1 found nothing

   • The first myostatin blocker tested in people was safe enough but did nothing for strength. 116 adults with three kinds of muscular dystrophy were split by chance across four dose levels or a dummy, and nobody involved knew which. Skin reactions appeared at the two highest doses. The authors state plainly that no improvement was seen in muscle strength or function. Muscle size crept up slightly on scans in a few people. The study was never built to prove the drug works, only to check it was tolerable.

     The numbers 116 adults with Becker, facioscapulohumeral or limb-girdle muscular dystrophy, randomised, double-blind, placebo-controlled, across four dose cohorts from 1 to 30 mg/kg. Safety and tolerability were good apart from cutaneous hypersensitivity at 10 and 30 mg/kg. Verbatim: there were no improvements noted in exploratory end points of muscle strength or function, though the study was not powered for efficacy. Bioactivity was supported by a trend toward increased muscle size on DXA and histology in a limited number of subjects.

     Annals of Neurology, 2008· 116 people

   • found nothingInside the muscle, individual fibres did look better, in a handful of people. Researchers took muscle fibre samples from trial participants before and after six months. Five of them had been given the myostatin blocker and one had the dummy. Four of those five showed improved force and speed in single fibres. The one on the dummy showed none. Six people is not a result, it is a hint. Whole-muscle strength in these same patients did not improve, so whatever happened at the cellular level did not reach them.

     The numbers 6 patients from the MYO-029 trial, 5 on drug and 1 on placebo, with single muscle fibres examined directly. Four of the five on drug showed improved single fibre force and speed and the placebo patient showed none. Whole muscle strength did not improve in those same patients, which is the result that matters to a person. The only look inside the muscle itself from this programme.

     Muscle and Nerve, 2009· 6 people

7. 07DasatinibSmall moleculeLimited. People have taken it in a published study, but fewer than 100 of them. 3 completed trials, 74 people given it.3 reported

   14 people given it2 trials

   • An early study in 14 people with lung scarring found their physical function improved. Everybody knew they were taking the drugs and there was no comparison group. Side effects were mostly mild to moderate. Read this as the study that justified running a proper trial, and then read what the proper trial went on to find.

     The numbers Open-label pilot in 14 IPF patients: improved physical function measures; adverse events mostly mild to moderate.

     EBioMedicine, 2019· 14 people

   • A pilot that looked positive lost its effect the moment a dummy group was added. The same team ran it again in twelve people with lung scarring. Frailty, lung function and physical function did not meaningfully separate. The people on the drugs reported 65 non-serious side effects against 22 on the dummy. Small numbers, and a clear lesson about how much to trust an uncontrolled pilot.

     The numbers The same team's open-label pilot repeated with a placebo arm, n=12: frailty, pulmonary and physical function did not appear to differ meaningfully between groups, and the drug arm reported 65 non-serious adverse events against 22 on placebo.

     EBioMedicine, 2023· number of people not given

   Also measured in animals or in the lab

   • In old mice: D+Q given every so often helped them move and function better. From the point treatment began, they lived about 36 percent longer.

     The numbers Intermittent D+Q in old mice improved physical function and extended remaining lifespan by about 36%.

     Nature Medicine, 2018· animal study

8. 08FollistatinBiologicLimited. People have taken it in a published study, but fewer than 100 of them. 2 completed trials, 12 people given it.

   12 people given it2 trials

   • A first small look at gene therapy for a muscle wasting disease, not an answer. Six men with Becker muscular dystrophy had the follistatin gene injected into their thigh muscles. Four of the six walked further in six minutes afterwards, by between 29 and 125 metres. Muscle samples showed less scarring. No harms were reported. Here is the problem. Everybody got the treatment, so there is nothing to compare against, and six people cannot tell you whether this works.

     The numbers AAV1-delivered FS344 injected into the quadriceps of six patients: four improved six-minute walk distance (29 to 125 m), no adverse effects, reduced fibrosis on biopsy.

     Molecular Therapy, 2015· 6 people

   • Six men with a progressive muscle disease walked further after gene therapy, while untreated men like them got worse. The treated group gained a typical 56 metres a year on a six minute walk. The comparison men lost 26 metres a year. Here is the problem. Nobody was picked by chance for either group, so the men who got the treatment may have differed from the men who did not in ways that explain the gap. Six people is far too few to settle it.

     The numbers Six treated patients gained a median 56 m/year on six-minute walk versus a 26 m/year decline in matched untreated controls; small, non-randomized.

     Molecular Therapy, 2017· 6 people

9. 09MK-677Small moleculeLimited. People have taken it in a published study, but fewer than 100 of them. 2 completed trials, 65 people given it.11 reported

   nobody was given it1 watch-only study

   • The trial was stopped early over a heart safety signal, and most of what it measured did not improve. 123 older people recovering from a broken hip took MK-677 or a dummy pill. Walking speed improved and IGF-1 rose. Most other measures of function did not move. The study was halted before it finished because of concern about congestive heart failure in treated patients. Stopping early leaves the benefit thin and the harm not properly measured.

     The numbers 123 elderly hip-fracture patients: gait speed improved and IGF-1 rose, but most functional measures did not, and the trial was terminated early over a congestive heart failure safety signal.

     Archives of Gerontology and Geriatrics, 2011· number of people not givenobserved only, nobody was given it

10. 10ACE-031BiologicLimited. People have taken it in a published study, but fewer than 100 of them. 1 completed trial, 48 people given it.5 reported

    nobody was given it1 watch-only study

    • The trial was stopped early for safety, and it never showed a benefit. Boys with Duchenne muscular dystrophy were given ACE-031 or a dummy injection. After the second dose level, the study was halted because boys were getting nosebleeds and small spidery blood vessels in the skin. Walking distance and muscle mass leaned in the drug's favour, but not by enough to call it a real effect. Treat this as a harm signal with no proven upside.

      The numbers The trial was stopped after the second dose level because of epistaxis and telangiectasias; only non-significant trends toward preserved six-minute walk and higher lean mass were seen.

      Muscle and Nerve, 2017· number of people not givenobserved only, nobody was given it

11. 11TestosteroneHormoneModerate. At least one finished study in people, and at least 100 people who took it. 1 completed trial, 5,246 people given it.10 reported

    nobody was given it1 watch-only study1 found nothing

    • found nothingTestosterone helped older men with sex and did much less for everything else. It raised their levels to what a man in his twenties or thirties would have. Sex drive, sexual activity and erections all clearly got better. Energy did not get better at all. Walking distance only improved when three studies were lumped together, which is a weaker way to find a result. Mood was slightly better. Side effects were no more common than on the dummy gel.

      The numbers The T-Trials. Treatment raised testosterone into the mid-normal range for men aged 19 to 40. Sexual activity, sexual desire and erectile function all improved significantly. Vitality did not improve on the fatigue scale. The proportion improving 6-minute walking distance by at least 50 m did not differ in the Physical Function Trial but did when all three trials were pooled, 20.5% against 12.6%, p = 0.003. Mood and depressive symptoms were slightly better. Adverse event rates were similar.

      New England Journal of Medicine, 2016· number of people not givenobserved only, nobody was given it

Other compounds

1. 01RuxolitinibSmall moleculeModerate. At least one finished study in people, and at least 100 people who took it. 1 completed trial, 309 people given it.

   1 animal or lab study

   Measured in animals or in the lab

   • In human cells in a dish and in old mice. Blocking the JAK pathway quieted the inflammatory signals from zombie cells. In old mice, 10 weeks of a JAK-blocking drug cut inflammation and helped them move better. There is no result in people.

     The numbers Human cells in culture plus aged mice. Senescent human preadipocytes and HUVECs developed a SASP that was suppressed by JAK-pathway RNAi or JAK inhibitors, and conditioned medium from JAK-inhibitor-treated senescent cells was much less proinflammatory. Giving a JAK inhibitor to aged mice for 10 weeks alleviated adipose and systemic inflammation and enhanced physical function. The authors state their findings are consistent with a possible contribution of senescent cells to age-related frailty and speculate about SASP inhibition; they do not report a human outcome.

     Proceedings of the National Academy of Sciences, 2015· animal study

2. 02Zoledronic acidSmall moleculeStrong. At least two finished studies in people, and at least 1,000 people who took it. 4 completed trials, 7,176 people given it.

   1 animal or lab study

   Measured in animals or in the lab

   • In human lung cells and mouse cells in a dish, and in old mice. Zoledronic acid killed zombie cells while barely touching healthy ones. In old mice, eight weeks of it cut inflammatory signals in the blood and improved grip strength. It does not test whether it helps people live longer.

     The numbers Mayo Clinic. In human lung fibroblasts and DNA-repair-deficient mouse embryonic fibroblasts, zoledronic acid killed senescent cells with minimal effect on non-senescent cells. In aged mice treated for eight weeks it significantly reduced circulating SASP factors including CCL7, IL-1 beta, TNFRSF1A and TGF beta 1, and improved grip strength. Cell and mouse work proposing a senolytic or senomorphic mechanism for the non-skeletal effects; it does not test the mortality question in people.

     Aging (Albany NY), 2023· animal study

3. 03GYM329BiologicPreliminary. Studied in people, but no published study has had people take it. 2 human studies where nobody was given it, 1 registered with no result.1 reported

   1 animal or lab study

   Measured in animals or in the lab

   • found nothingIn mice. In mouse models of muscle disease, this antibody made muscles stronger. That matters because strength is exactly what the three earlier antibodies failed to improve in people. The earlier antibodies also looked hopeful in animals first.

     The numbers The preclinical characterisation. Mouse disease models. Reports enhanced muscle strength, which is notable precisely because that is the endpoint the three earlier antibodies failed to move in humans. Animal work, and the site keeps that separate from human evidence for exactly this reason.

     Scientific Reports, 2021· animal study