Rapamycin Plus Resistance Training: The First Randomized Trial Found No Benefit
The first randomized trial of weekly rapamycin plus exercise found no functional benefit in older adults. What RAPA-EX-01 measured and what to track.
Rapamycin Plus Resistance Training: The First Randomized Trial Found No Benefit **As of 2026, the only randomized controlled trial to test weekly low-dose rapamycin alongside a structured exercise program found no added benefit, and every functional outcome leaned toward placebo. RAPA-EX-01, published in April 2026 in the Journal of Cachexia, Sarcopenia and Muscle (2026;17(2):e70274), randomized 40 sedentary adults aged 65 to 85 to sirolimus 6 mg once weekly or matched placebo for 13 weeks while both arms performed the same home-based chair-stand and exercycle program three times weekly. The primary intention-to-treat result on the 30-second chair-stand test was an adjusted mean difference of -2.13 repetitions (95% CI -4.61 to 0.34; p=0.089), directionally favoring placebo, and two prespecified sensitivity analyses reached statistical significance in the same direction. Importantly, the trial measured functional capacity, not muscle size, so it cannot say whether rapamycin blunted hypertrophy specifically. The authors describe it as an exploratory trial, it enrolled 40 previously sedentary older adults for 13 weeks rather than trained lifters, and any decision about rapamycin belongs with a licensed prescriber.** For years, the rapamycin-and-exercise question has been argued from mouse data and mechanism diagrams. In April 2026 it finally got a human randomized controlled trial. RAPA-EX-01 is the first study to give people weekly low-dose rapamycin and a standardized training program at the same time, under double-blind, placebo-controlled conditions, and to measure what happened to their physical function. The result was not what the hypothesis predicted. This article walks through what the trial was designed to test, what it found, what it explicitly could not measure, and how someone tracking their own health data, or a coach tracking a client's, can organize the relevant numbers. MyProtocolStack is a tracking and education tool. Nothing here is medical advice, and nothing here is a recommendation to start, stop, or change any medication or training program.
What RAPA-EX-01 Tested, and Why the Design Matters
RAPA-EX-01 was a randomized, double-blind, placebo-controlled trial of 40 sedentary adults aged 65 to 85, with a mean age of 72.2 years and 47.5% female participants. Enrollees were assigned 1:1 to either sirolimus 6 mg once weekly or a matched placebo for 13 weeks. Sirolimus is the generic name for rapamycin. The trial ran at a single site, the Aotearoa Clinical Trials Trust in Auckland, New Zealand, was registered as ACTRN12624000790549, and was funded by public crowdfunding facilitated by Lifespan.io and VitaDAO. It was published by Stanfield, Leroux, Kaeberlein, Jones, and Lucas in the Journal of Cachexia, Sarcopenia and Muscle, 2026;17(2):e70274. The authors label it an exploratory trial, which is the correct weight to give it.
The design detail that makes this trial useful is that both arms exercised. Every participant followed the same standardized home-based program three times weekly: chair-stands for the resistance component and a stationary exercycle for the endurance component. That means the comparison is not drug versus nothing. It is exercise plus drug versus exercise plus placebo, which is the question people following this space actually ask.
Two further design choices matter. First, the primary outcome was the 30-second chair-stand test, a validated functional measure of lower-body strength and endurance that counts how many times a person can rise from a chair in half a minute. Second, the dosing was deliberately timed relative to training. Participants trained on days 1, 3, and 5 of each week and took the study drug on day 6, roughly 24 hours after the final session of the week. That timing was not incidental. It was meant to give the hypothesis its best possible shot.
The Cycling Hypothesis: The Idea Being Tested
The trial existed to test what researchers call the cycling hypothesis. The reasoning goes like this. mTORC1, the growth-signaling complex that rapamycin inhibits, is also the pathway that resistance training activates to drive muscle adaptation. Preclinical work suggested that alternating between inhibition and activation, rather than holding the pathway suppressed continuously, might improve adaptation to exercise by separating a catabolic and autophagic window driven by the drug from an anabolic window driven by training recovery.
There was a plausible aging-specific angle too. Older muscle can show elevated basal mTORC1 signaling alongside a blunted protein-synthesis response to training, a phenomenon described as anabolic resistance. If chronically elevated baseline signaling is part of what dulls the training response, then intermittently lowering it might restore sensitivity.
That is a genuinely interesting hypothesis. It is also, critically, a preclinical hypothesis. The trial's own abstract opens by attributing the cycling idea to preclinical models and states plainly that whether it translates to older adults was unknown. It came from animal and mechanistic work, not from human trials. RAPA-EX-01 is the first attempt to test it in people under randomized, controlled, blinded conditions, and this is exactly the distinction that gets lost when mechanism is discussed as though it were evidence. The animal and cell data motivated the study. The study is what tells us whether the idea holds up in humans.
The Results: Every Functional Outcome Leaned Toward Placebo
The headline result is that once-weekly sirolimus did not enhance functional gains from the exercise program, and in the sensitivity analyses it appeared to modestly attenuate them. Both groups improved their chair-stand performance over the 13 weeks. The drug arm improved less.
|---|---|---|---|---|
A negative number in that table means the sirolimus arm did worse than placebo. The primary intention-to-treat analysis did not reach statistical significance at p=0.089, which in a strict reading means the primary endpoint was null. But the two prespecified sensitivity analyses, complete-case at p=0.045 and per-protocol at p=0.007, both crossed the conventional threshold and both pointed the same way.
Sample sizes shrink across those analyses, which is worth knowing when you weigh them. Forty people were randomized, 20 per arm. Five participants discontinued in the sirolimus arm and three in the placebo arm. The complete-case analysis included 16 sirolimus and 19 placebo participants, and the per-protocol analysis included 15 and 16.
That pattern deserves an honest reading in both directions. Sensitivity analyses that restrict to participants who completed the trial or adhered to protocol are less protected against bias than intention-to-treat, so they carry less weight on their own, and the smaller the surviving sample the more caution is warranted. At the same time, when the primary analysis is directionally negative and every prespecified sensitivity analysis and every secondary outcome points the same way, the consistency itself is informative. There is no outcome in this trial where the drug arm came out ahead. What the trial did not deliver is any support for the cycling hypothesis.
What the Safety and Laboratory Data Showed
Tolerability was not dramatic, but it was not neutral either. Seventeen participants in each arm, 85%, reported at least one adverse event, so the proportion affected was identical. The total burden differed: 99 events on sirolimus versus 63 on placebo. There was one serious adverse event judged possibly drug-related, a case of pneumonia. Given that rapamycin is an immunosuppressant, an infection signal in a 40-person trial is worth noting even though a single event cannot establish a rate.
The published protocol specified a broad laboratory panel at screening and end of study, including full blood count, urea and electrolytes, estimated glomerular filtration rate, liver function tests, [HbA1c](/biomarkers/hba1c), lipids, serum [IGF-1](/biomarkers/igf-1), and [hs-CRP](/biomarkers/hs-crp), with an interim safety draw at week 6.
Several of those markers moved. The adjusted between-group differences reported in the trial include HbA1c higher by 1.74 mmol/mol in the sirolimus arm (p=0.030) and [LDL cholesterol](/biomarkers/ldl-c) higher by 0.32 mmol/L (p=0.036). Both of those reached statistical significance, so this is not merely a directional wobble. Mean corpuscular volume fell by 2.90 fL (p<0.001), platelet count rose by 17.6 x10^9/L (p=0.025), and alkaline phosphatase rose by 5.56 U/L (p=0.012). C-reactive protein was higher in the sirolimus arm by 4.28 mg/L but did not reach significance (95% CI -0.04 to 8.68; p=0.15), and that estimate was heavily influenced by two outlier values. The glucose and lipid movement mirrors what is already documented about mTOR inhibition and metabolic handling. These are group-level averages over 13 weeks in 40 people, not a prediction about any individual.
On the epigenetic clocks, no measure reached statistical significance. PCGrimAge trended toward a younger biological age in the sirolimus arm by 2.28 years (p=0.098), while SystemsAge was 1.04 years higher (p=0.24), OMICmAge 0.28 years higher (p=0.592), and DunedinPACE essentially unchanged (p=0.905). That is a null result on biological age, and it is a useful reminder of why [biological age markers](/blog/biological-age-markers) should be read as research instruments rather than scoreboards.
Where This Sits Alongside the PEARL Trial
This trial does not quietly overturn our earlier coverage of the [PEARL trial](/blog/rapamycin-pearl-trial-mtor-longevity-2026), and it is worth being explicit about why, because the two results are easy to mash together incorrectly.
PEARL randomized 114 healthy adults aged 50 to 85 to weekly 5 mg, weekly 10 mg, or placebo over 48 weeks. Its own primary endpoint, visceral adiposity by DXA, did not change significantly. The finding that drew attention was a secondary one: women on the 10 mg weekly dose gained about 6% more lean tissue mass than placebo (mean difference 6.19%; p=0.018) and reported less pain, with adverse-event rates similar to placebo. RAPA-EX-01 is a different question in a different population over a different timeframe. It ran 13 weeks, not 48. It enrolled sedentary 65 to 85 year olds, not a broad healthy 50-plus cohort. Every participant trained, which PEARL did not require. And most importantly, RAPA-EX-01 measured function, not body composition. There was no DXA scan and no muscle biopsy in this trial, a limitation the authors state directly.
That last point is the honest answer to the target question. RAPA-EX-01 cannot tell you whether rapamycin blunted muscle growth, because it never measured muscle. What it can tell you is that in this population, over this window, the people taking rapamycin got less functional improvement from the same training program than the people taking placebo. Those are related but distinct claims, and conflating them would misstate the evidence.
The authors flag a second constraint that matters for anyone reading this as a lifting question. The home-based chair-stand protocol used body weight rather than external load, so it has a lower ceiling for maximal strength development than heavy resistance training. A program that never approaches a heavy stimulus is not a clean test of whether the drug interferes with a heavy stimulus. Both trials remain small, short, and preliminary, and neither settles the human longevity question.
What to Track: Biomarkers and Functional Tests Worth Logging
If you or your provider are following this research, the useful move is to keep a clean longitudinal record rather than react to a single study. The list below reflects what RAPA-EX-01 and the surrounding literature actually measured. It is not a recommendation to test, dose, or change anything.
That last bullet is the one most people overlook. Blood markers require a draw and a wait. Functional tests cost nothing and can be repeated monthly under identical conditions, which is what makes them trackable. Pair them with [body composition tracking](/blog/how-to-track-body-composition) if you have access to a consistent method, and follow the same discipline you would use when [comparing lab results over time](/blog/how-to-compare-lab-results-over-time): same conditions, same protocol, same time of day, every time.
[Log your functional tests and lab trends side by side with MyProtocolStack.](/auth/login?mode=signup)
For Coaches: A Measurement Protocol, Not a Recommendation
This is the first controlled dataset a coach can point to when a client is on weekly rapamycin under a prescriber's care. It changes what you can reasonably say and, more usefully, what you can reasonably measure.
What it does not do is give you grounds to advise anything about the medication. Rapamycin is a prescription immunosuppressant, any longevity use is off-label, and dosing decisions sit with the prescribing clinician. A coach who tells a client to skip a dose before a training block has crossed a line that no trial result justifies.
What you can do is instrument the client's own data. A defensible protocol looks like this:
Framed that way, you are not practicing medicine and you are not second-guessing a prescription. You are producing the one thing a busy clinician almost never gets: consistent, longitudinal functional data from between visits. The same evidence-reading discipline applies here as anywhere else, and it is worth revisiting how to [evaluate research claims](/blog/how-to-evaluate-peptide-research) before letting any single trial reshape a program.
The Practical Takeaway for 2026
RAPA-EX-01 answered a narrow question cleanly. In 40 sedentary older adults over 13 weeks, weekly 6 mg sirolimus did not improve functional gains from a home exercise program, the primary intention-to-treat result was null but directionally negative, both prespecified sensitivity analyses were significantly negative, every secondary outcome pointed the same way, adverse-event burden was higher on the drug, and HbA1c and LDL-C both moved significantly in the unfavorable direction.
What it did not answer is nearly as important. It did not measure muscle mass or hypertrophy, because it ran no DXA scans and no biopsies. It did not use heavy external load, so it is not a clean test of the heavy-lifting scenario people usually have in mind. It did not enroll trained younger lifters, so it says nothing about how a well-trained 35-year-old would respond. It ran three months, so it says nothing about multi-year effects in either direction. And it was not designed to address whether rapamycin influences the diseases that actually drive mortality. The authors' own conclusion is that longer or less frequent dosing regimens still need testing before anyone can judge the benefit-risk balance. On the specific question of whether stacking weekly rapamycin on top of a training program makes the training work better, the first randomized evidence says no.
For anyone tracking their own health, the constructive response is not to chase or dismiss a headline. It is to keep records that stay useful regardless of how the next trial lands: consistent functional tests, consistent lab conditions, dates attached to everything, and a trend line ready for the person qualified to interpret it. You can browse plain-language explanations of each test in the full [biomarker library](/biomarkers).
Frequently Asked Questions
Does rapamycin blunt muscle gains from exercise?
The first randomized controlled trial to test this, RAPA-EX-01, found that weekly 6 mg sirolimus did not enhance functional gains from a 13-week home exercise program in 40 sedentary adults aged 65 to 85, and in prespecified sensitivity analyses it appeared to modestly attenuate them. An important caveat is that the trial measured physical function, specifically chair-stands, 6-minute walk distance, and grip strength, and included no DXA scans and no muscle biopsies. So it cannot confirm that muscle growth specifically was blunted, only that functional improvement was smaller in the drug arm. The exercise program also used body weight rather than heavy external load, which the authors note has a lower ceiling for maximal strength development. This was one small, short, exploratory trial in previously sedentary older adults, and any decision about rapamycin belongs with a licensed prescriber.
What exactly did the RAPA-EX-01 trial find?
RAPA-EX-01, published in April 2026 in the Journal of Cachexia, Sarcopenia and Muscle 2026;17(2):e70274, randomized 40 sedentary adults aged 65 to 85 to sirolimus 6 mg weekly or placebo for 13 weeks while both arms did the same thrice-weekly chair-stand and exercycle program. The primary intention-to-treat result on the 30-second chair-stand was an adjusted mean difference of -2.13 repetitions favoring placebo (95% CI -4.61 to 0.34; p=0.089), which was not statistically significant. Prespecified sensitivity analyses were significant and also favored placebo: complete-case -2.46 reps (p=0.045) and per-protocol -3.44 reps (p=0.007). Secondary outcomes including 6-minute walk distance, grip strength, and SF-36 all trended toward placebo without reaching significance, and the sirolimus arm recorded 99 adverse events versus 63 on placebo.
Why did the sensitivity analyses reach significance when the primary analysis did not?
Intention-to-treat analysis includes everyone as randomized, which protects against bias but can dilute an effect if participants drop out or do not adhere. Complete-case and per-protocol analyses restrict to participants who finished or who followed the protocol, which can sharpen the estimate but also removes some of that protection, so they carry less weight on their own. In RAPA-EX-01 the primary intention-to-treat result was null at p=0.089 while both prespecified sensitivity analyses crossed significance in the same direction, on smaller samples of 35 and 31 participants respectively. The honest reading is that the primary endpoint was formally negative, and that the consistency across every analysis and every secondary outcome, none of which favored the drug, is itself informative.
Did rapamycin change any blood markers in this trial?
Yes, several. Compared with placebo, the sirolimus arm showed a higher HbA1c by 1.74 mmol/mol (p=0.030) and higher LDL cholesterol by 0.32 mmol/L (p=0.036), both statistically significant. Mean corpuscular volume fell by 2.90 fL (p<0.001), platelet count rose by 17.6 x10^9/L (p=0.025), and alkaline phosphatase rose by 5.56 U/L (p=0.012). C-reactive protein was higher by 4.28 mg/L but did not reach significance (p=0.15) and was skewed by two outliers. Epigenetic age measures including PCGrimAge, SystemsAge, OMICmAge, and DunedinPACE showed no statistically significant differences. These are 13-week group averages in 40 people and are not a forecast for any individual.
Does this contradict the PEARL trial's lean-mass finding?
Not directly, and the two should not be merged. PEARL randomized 114 healthy adults aged 50 to 85 over 48 weeks. Its primary endpoint, visceral adiposity by DXA, was null, and the lean-mass result was a secondary finding: women on 10 mg weekly gained about 6% more lean tissue mass than placebo (mean difference 6.19%; p=0.018), with adverse-event rates similar to placebo. RAPA-EX-01 ran 13 weeks in 40 sedentary 65 to 85 year olds who all exercised, used a different dose, and measured physical function rather than body composition. Because RAPA-EX-01 included no DXA scan or biopsy, it can neither confirm nor refute PEARL's lean-mass result. Both trials are small, short, and preliminary, and neither settles the human longevity question.
What was the cycling hypothesis, and did the trial support it?
The cycling hypothesis is the preclinical idea that alternating inhibition and activation of mTORC1, rather than suppressing it continuously, could enhance adaptation to exercise by separating a drug-driven catabolic and autophagic window from a training-driven anabolic window. RAPA-EX-01 was designed to test it in humans, and dosing was deliberately scheduled on day 6 of each week, roughly 24 hours after the last of three weekly sessions, to give the hypothesis its best chance. The trial produced no support for it. It is worth being clear that the cycling hypothesis came from animal and mechanistic research, not from human trials, and this is the first randomized human test of the idea.
Sources
1. Stanfield B, Leroux B, Kaeberlein M, Jones J, Lucas R. "Exercise and Weekly Sirolimus (Rapamycin) in Older Adults: RAPA-EX-01 Randomised, Double-Blind, Placebo-Controlled Trial." Journal of Cachexia, Sarcopenia and Muscle. 2026;17(2):e70274. PMID 41985884. https://doi.org/10.1002/jcsm.70274
2. Free full text of the RAPA-EX-01 trial, PubMed Central PMC13082878. https://pmc.ncbi.nlm.nih.gov/articles/PMC13082878/
3. PubMed record and abstract, RAPA-EX-01 trial, PMID 41985884. https://pubmed.ncbi.nlm.nih.gov/41985884/
4. Stanfield B, Kaeberlein M, Leroux B, Jones J, Lucas R, Arroll B. Published RAPA-EX-01 trial protocol. Trials, 2024. PMC11443903. https://pmc.ncbi.nlm.nih.gov/articles/PMC11443903/
5. Moel M, Harinath G, Lee V, Nyquist A, Morgan SL, Isman A, Zalzala S. "Influence of rapamycin on safety and healthspan metrics after one year: PEARL trial results." Aging (Albany NY). 2025;17(4):908-936. PMID 40188830. https://doi.org/10.18632/aging.206235
6. Peter Attia MD, "Disappointing results from the first rapamycin-plus-exercise trial." https://peterattiamd.com/rapamycin-plus-exercise-trial/
*MyProtocolStack is a tracking and education tool, not medical advice, diagnosis, or treatment, and you should always consult a qualified healthcare professional before making any changes to your health protocol.*
Enter your blood work in MyProtocolStack, run StackAI analysis, and get personalized insights based on your actual numbers — not generic charts.
Start Free →