AICAR carries one of the most quoted numbers in the "exercise pill" story: mice that did no training ran 44% longer after four weeks on it. That 2008 mouse result put a drug that had already failed elsewhere into the hands of endurance athletes. For anyone tested, the twist is that the body makes AICAR itself, so the test has to tell the difference between what a person produced and what they took. This page covers status, the human record and how laboratories catch it. It gives no usage guidance, per our editorial standards.
What it is, in one paragraph
AICAR stands for 5-aminoimidazole-4-carboxamide ribonucleoside. It is a small molecule, not a peptide: formula C9H14N4O5, molecular weight 258.23, PubChem CID 17513, where it is filed under its international drug name, acadesine. Inside cells it is converted to ZMP, which mimics AMP and switches on AMPK, the enzyme that senses low energy and pushes muscle toward burning fat, the same switch endurance training flips. One naming trap for anyone checking: a PubChem lookup of the bare string "AICAR" resolves to the phosphorylated form, the ribotide (CID 65110, molecular weight 338.21), not the riboside sold and studied under that name. This site's endurance page places AICAR in the wider exercise-mimetic market; its stablemate in the same WADA line is MOTS-c.
Status: named, at all times, non-Specified
WADA's 2026 Prohibited List, read on 2026-10-03, puts AICAR in S4.4.1 within the metabolic modulators:
"Activators of the AMP-activated protein kinase (AMPK), e.g. … (BAM15), AICAR, mitochondrial open reading frame of the 12S rRNA-c (MOTS-c)"
The same section continues with PPARδ agonists such as GW1516 and Rev-erbα agonists such as SR9009. The List states that classes S4.3 and S4.4 are non-Specified, and S4 applies at all times. The rest of the map is on WADA status by compound.
The sanctions record
USADA's published sanctions table held 1,075 entries on 2026-10-03. One names AICAR:
| Announced | Sport | Violation, as published | Sanction |
|---|---|---|---|
| 2017-12-08 | Track and field | Non-analytical: use and possession (AICAR) | 4-year suspension; loss of results |
The case was non-analytical: established by evidence other than a positive sample. For comparison, GW1516, its neighbour in the same section, appears in 22 USADA sanctions. AICAR's single entry reflects, in part, how hard it is to catch on concentration alone (below).
The human record: five registrations, one big failure
ClinicalTrials.gov returns 2 registrations for "AICAR" and 3 for "acadesine" as an intervention (2026-10-03). Read one by one:
| Registration | What it was | Size | Status |
|---|---|---|---|
| NCT00872001, RED-CABG | acadesine to protect the heart during bypass surgery, phase 3 | 3,080 randomised | terminated for futility |
| NCT00559624 | acadesine in B-cell chronic lymphocytic leukaemia, phase 1/2 | 40 | completed |
| NCT01813838 | acadesine, phase 1/2 | 5 | terminated |
| NCT00004314 | AICAR in Lesch-Nyhan disease, phase 2 pilot | 2 | completed |
| NCT00168519 | exercise-mediated glucose uptake in type 2 diabetes; AICAR one of five agents in a single arm | 40 | completed |
None tests athletic performance, endurance or muscle in healthy people.
The large one is worth reading. RED-CABG randomised intermediate- to high-risk patients at 300 sites in seven countries to acadesine infused at 0.1 mg/kg per minute for 7 hours or placebo. A pre-specified futility analysis stopped it after 3,080 of a projected 7,500 patients. The primary outcome (death, stroke or severe heart failure needing mechanical support by day 28) occurred in 76 of 1,493 (5.1%) on acadesine and 75 of 1,493 (5.0%) on placebo, odds ratio 1.01 (Newman 2012, JAMA). Drugs@FDA has no application under either name.
What the performance claim rests on
The claim rests on one experiment. In a 2008 Cell paper, sedentary adult mice given AICAR for four weeks showed induced metabolic genes and 44% greater treadmill endurance, with no training at all (Narkar 2008). The paper's own framing was an "exercise mimetic" for metabolic disease, not a sports supplement. Eighteen years later no human trial has measured performance, endurance, body composition or recovery with it in athletes. What people did get, in the largest trial, was no measurable benefit on the outcome it was designed to improve.
Can a test find it? The endogenous problem
AICAR is an intermediate in the body's own purine synthesis, so every urine sample contains some. That makes it a concentration-and-origin problem, like testosterone, rather than a yes-or-no detection.
Step 1, concentration. Two reference populations have been published, and they disagree:
| Study | Samples | Reported level |
|---|---|---|
| Thomas 2010, Cologne | 499 athletes' doping-control samples | mean 2,186 ng/mL (SD 1,655) |
| Sobolevsky 2019, Los Angeles | 12,377 US athlete samples | mean 647 ng/mL, median 574, 99th percentile 1,786, 99.7th 2,151 |
The German mean is 3.4 times the US one. The 2010 paper notes that concentrations differed by sex, sport and in- versus out-of-competition sampling; the two studies also used different methods and populations, so the gap is a reason no single fixed cut-off is quoted on this page. The US authors suggested that a sample above 2,000 or 2,500 ng/mL go forward to step 2.
Step 2, origin. Carbon isotope ratio mass spectrometry compares the carbon-13/carbon-12 ratio of the sample's AICAR with that of endogenous reference steroids. In a reference population of 63 men and women and an elimination study, the method detected a single oral AICAR dose for more than 40 hours (Piper 2014). A French laboratory published its working version in 2017 (Buisson 2017), and a 2022 paper proposed the ratio of AICAR to a related metabolite, SAICAr, as an additional marker (Sobolevsky 2022). Workplace drug tests do not look for it. More on how sport testing differs is on peptides and drug testing.
The honest summary
- Named on WADA's 2026 List in S4.4.1 (AMPK activators); prohibited at all times; non-Specified.
- 1 of 1,075 USADA published sanctions (2017, 4 years, non-analytical).
- Mouse endurance +44% (2008); no human performance trial.
- As acadesine: RED-CABG stopped for futility at 3,080 patients, 5.1% vs 5.0%, OR 1.01; never approved.
- Made by the body: urine reference levels of 574 ng/mL median (US) vs 2,186 ng/mL mean (Germany); origin settled by carbon isotope ratio, single oral dose visible >40 hours.
Limits of this page
The registry read covered what two intervention searches returned on 2026-10-03; trials under code names would be missed. The human pharmacology studies that used AICAR as a laboratory tool in clamp experiments were not counted as trials of the substance. USADA's table covers one agency's published sanctions, summarised in a few words each. The two urine reference studies differ in method and population, and their numbers are reported as published, not reconciled.
Sources and dates
- World Anti-Doping Agency, 2026 Prohibited List, sections S4 and S4.4.1 read 2026-10-03.
- U.S. Anti-Doping Agency, sanctions table, read 2026-10-03: 1,075 entries; one naming AICAR; 22 naming GW1516.
- Narkar VA, Downes M, Yu RT, et al. AMPK and PPARδ agonists are exercise mimetics. Cell 2008;134:405–15 — PMID 18674809
- Newman MF, Ferguson TB, White JA, et al. Effect of adenosine-regulating agent acadesine on morbidity and mortality associated with coronary artery bypass grafting: the RED-CABG randomized controlled trial. JAMA 2012;308:157–64 — PMID 22782417
- Thomas A, Beuck S, Eickhoff JC, et al. Quantification of urinary AICAR concentrations as a matter of doping controls. Anal Bioanal Chem 2010;396:2899–908 — PMID 20225061
- Piper T, Thomas A, Baume N, et al. Determination of 13C/12C ratios of endogenous urinary AICAR. Rapid Commun Mass Spectrom 2014;28:1194–202 — PMID 24760559
- Buisson C, Frelat C, Mongongu C, Martinat N, Audran M. Implementation of AICAR analysis by GC-C-IRMS for anti-doping purposes. Drug Test Anal 2017;9:1704–12 — PMID 29032594
- Sobolevsky T, Ahrens B. Urinary concentrations of AICAR and mannitol in athlete population. Drug Test Anal 2019;11:530–5 — PMID 30548818
- Sobolevsky T, et al. AICAr to SAICAr ratio can serve as additional marker of AICAr use. Drug Test Anal 2022;14:2017–25 — PMID 36342242
- ClinicalTrials.gov API v2, intervention searches "AICAR" (2) and "acadesine" (3), each record read, 2026-10-03. PubChem CID 17513 (acadesine; synonyms include "AICAR (Acadesine)") and CID 65110 (returned for the name "AICAR"; the ribotide), read 2026-10-03. openFDA Drugs@FDA: no match for acadesine or AICAR, 2026-10-03.