AICAR (Acadesin)

What it is

AICAR is not an invention of the peptide scene but a molecule that every human makes: an intermediate step in the synthesis of purines, measurable in the blood and urine of healthy people. Reproduced synthetically as an active substance, it is called acadesin — and unlike most substances in this corner, it did not get stuck in animal experiments but went through an entire development chain up to 2012. Only it was never about endurance.

The trick is called ZMP

AMPK is the cell’s energy guardian: when ATP falls and AMP rises, it switches to less building and more burning.

AICAR takes this route via a forgery. In the cell it is phosphorylated to its monophosphate ZMP, and to AMPK, ZMP looks like AMP. The 1995 study that showed this in rat liver cells worked out the decisive point: ZMP mimics both activating effects of AMP without ATP, ADP or AMP shifting. The cell believes in an energy shortage that does not exist. The catch: ZMP does not talk only to AMPK; a 2021 review compiled the AMPK-independent effects.

The 2008 mouse study

The sentence that made AICAR famous is in the title of a paper from Ronald Evans’s lab at the Salk Institute: AMPK and PPARdelta agonists are exercise mimetics. The PPARdelta agonist GW1516, better known as GW501516, achieved nothing at all on its own; only with training did running time rise by 68 percent and distance by 70 percent. AICAR was the surprise: four weeks of treatment increased running performance in untrained mice by 44 percent.

One detail is almost always swallowed: in the abstract AICAR is called “orally active”. But it was administered intraperitoneally, that is, injected into the abdominal cavity. Whether it works when swallowed was not tested there, and in humans the picture looks different.

What has been measured in humans

In a placebo-controlled double-blind study in healthy men, the terminal plasma half-life was 1.4 hours and bioavailability after intake in solution was below 5 percent. Six hours after an infusion, uric acid accounted for all of the labeled material in the plasma.

Two controlled studies measured metabolic effects. In 29 healthy men, glucose uptake in muscle rose 2.1-fold after three hours of AICAR, and 4.7-fold during cycling; in 10 men with type 2 diabetes, an infusion lowered glucose output by the liver. In both studies, AMPK phosphorylation in muscle did not rise measurably at all: the substance does something, but not demonstrably via the switch it is named after.

The clinical route as acadesin

The great hope lay in the heart. During bypass surgery, the heart muscle is temporarily cut off from its blood supply and then reconnected, a model case for damage from oxygen deprivation and reperfusion. Acadesin was supposed to protect the tissue, given as a seven-hour infusion. A 1997 meta-analysis of five trials with 4043 patients came out positive: perioperative heart attacks minus 27 percent.

Then came the counter-test. The RED-CABG trial tested the same infusion in high-risk patients at 300 centers, double-blind, with a hard endpoint up to day 28. It was stopped after 3080 of a planned 7500 patients: 5.0 percent under placebo, 5.1 percent under acadesin. The second strand was oncology, with a phase I/II study in 24 patients with chronic lymphocytic leukemia; there were initial signs of efficacy, but they were inconsistent.

Doping and the detection problem

AICAR has been banned in elite sport since 2009; in the 2026 Prohibited List it is named among the AMPK activators, alongside MOTS-c. Testing is difficult because the body makes AICAR itself. In 499 doping control samples, the mean urine concentration was 2,186 ng/ml, with a standard deviation of 1,655 ng/ml — a spread that blunts threshold values. The way out is the carbon isotope ratio, with which a single dose was detectable for more than 40 hours.

In cycling, the substance has been an issue for years. In 2010 the French Anti-Doping Agency retrospectively suspected participants in the 2009 Tour de France, and a 2025 documentary by the German broadcaster ARD quotes an anonymous whistleblower saying AICAR is the drug of choice in the peloton — journalistic accounts, not laboratory findings. A doping expert reports there that he found 166 AICAR-like products, of which only 4 are on the Prohibited List.

How it differs from its neighbors

MOTS-c is an endogenous peptide that also acts via AMPK and rises with training in humans; administration from outside has only been tested in animals. SLU-PP-332 acts on the ERR receptors and has never been studied in humans. Metformin is the approved counterpart. No other substance in this group has seen as many humans as AICAR.

What the studies show

Narkar et al., Cell 2008 — the original study

Animal study in male C57Bl/6J mice. Untrained animals received AICAR for four weeks, intraperitoneally; the primary endpoint was running performance. Result: about 23 percent longer running time and about 44 percent longer distance, in 15 to 20 animals, endpoint met. Limitations: animal experiment, injection, treadmill as a surrogate.

Mangano, JAMA 1997 — meta-analysis of five trials

Individual patient data from five randomized, double-blind trials with 4043 patients after bypass surgery. Primary endpoint perioperative heart attack: odds ratio 0.69 with a confidence interval of 0.51 to 0.95, met. Limitation: the largest single trial in it had missed its prespecified endpoint.

Newman et al., JAMA 2012 — RED-CABG

Randomized, double-blind, placebo-controlled, 300 centers, high-risk patients before planned bypass surgery, mean age 66 years. Primary endpoint: death, stroke or mechanical support up to day 28. Stopped after 3080 of 7500 planned patients, 5.0 versus 5.1 percent, odds ratio 1.01, endpoint not met.

Cuthbertson et al., Diabetes 2007 — glucose uptake in humans

Controlled human study in 29 healthy men. After three hours, uptake of labeled glucose into muscle rose 2.1-fold under AICAR and 4.7-fold under cycling exercise, without an increase in AMPK activity. Surrogate marker, not a performance measure.

Where the data stop

On the question everyone is interested in, there is nothing: no human study has tested endurance, fat mass or body composition under AICAR. The controlled metabolic studies in humans measure hours, not weeks, and use surrogate markers instead of performance. The oral route is practically closed in humans, which is why the mouse figure of 44 percent cannot be transferred to a capsule. It is also uncertain what the effect should be attributed to: in both human studies, metabolism shifted without AMPK phosphorylation rising. On the hard clinical endpoint, the picture even flipped: what the meta-analysis showed in 4043 patients was not confirmed in the larger RED-CABG trial. On long-term safety, nothing exists.

Status, approval and legal

AICAR, or acadesin, is not an approved medicine and not a food supplement in Germany, the EU or the US. In the EU there are two orphan designations, EU/3/05/280 of May 27, 2005 and EU/3/11/881 of August 5, 2011 — this is not an approval but a support status. In sport, AICAR is named in the 2026 Prohibited List in section S4.4.1 among the AMPK activators; the ban applies at all times, and a therapeutic use exemption is not possible. The annex to the German Anti-Doping Act (Anti-Doping-Gesetz) lists AICAR by name; under § 2, trading, supplying and prescribing it for the purpose of doping are prohibited, as are acquisition and possession in non-small quantities.

Safety

The most important issue comes directly from the metabolic pathway: uric acid is the main metabolite in humans. In the heart studies, transient rises in serum uric acid occurred; in the leukemia study, hyperuricemia was so common that it was treated with preventive allopurinol. For people with gout or uric acid stones, this is the heart of the matter. In addition there are kidney function disorders, transient anemia and thrombocytopenia, and a drop in blood pressure during the infusion; in a study in MDS and AML, severe renal adverse effects led to the study being stopped. Acadesin also inhibits platelet aggregation; combinations with blood thinners or blood sugar–lowering drugs have not been studied. There is no upper limit from EFSA, BfR or NIH, because AICAR is not a nutrient. Anyone who wants to treat metabolic or heart problems has tested options, and these belong in a physician’s hands.

Sources

• Narkar et al., Cell 2008 — AMPK and PPARdelta agonists as exercise mimetics
• Corton et al., European Journal of Biochemistry 1995 — AICAR and ZMP as a tool for AMPK activation
• Dixon et al., Journal of Clinical Pharmacology 1991 — safety, tolerability and pharmacokinetics in humans
• Dixon et al., Journal of Clinical Pharmacology 1993 — distribution and metabolism, uric acid as the main metabolite
• Cuthbertson et al., Diabetes 2007 — AICAR increases glucose uptake in the muscle of healthy men
• Boon et al., Diabetologia 2008 — intravenous AICAR in type 2 diabetes
• Mangano, JAMA 1997 — meta-analysis of five randomized acadesin trials
• Newman et al., JAMA 2012 — RED-CABG, phase III, stopped for futility
• Van Den Neste et al., Cancer Chemotherapy and Pharmacology 2013 — phase I/II in chronic lymphocytic leukemia
• Cluzeau et al., International Journal of Molecular Sciences 2019 — acadesin in MDS and AML, phase I/II stopped because of severe renal adverse effects
• Višnjić et al., Cells 2021 — systematic review of AMPK-independent effects of AICAR