AICAR Isn't a Peptide, and Its Only Human Trials Failed

AICAR gets sold next to BPC-157 and TB-500 as another injectable research peptide for endurance and fat loss. It isn’t a peptide at all: it’s a ribonucleoside, a nucleotide-type molecule built on a purine base, with no amino acid chain in it anywhere^[1]. That alone would just be a labeling error, except the mislabeling hides something stranger underneath it.

AICAR has already been through the kind of randomized, placebo-controlled human trials that most peptide vendors can only gesture at: more than 4,000 patients, pooled across five international studies^[2]. None of those trials measured endurance, fat loss, or anything a gym-goer would recognize. They tested whether the drug could keep heart surgery patients from dying on the table. The largest and most rigorous of them found that it couldn’t^[3].

What AICAR’s human trials tested

Clinically, AICAR is called acadesine, and its entire human trial record runs through one procedure: coronary artery bypass graft (CABG) surgery.

• The pooled evidence base. A meta-analysis combining individual patient data from five randomized, double-blind, placebo-controlled trials across 81 centers in the US, Canada and Europe covered 4,043 CABG patients, split roughly evenly between acadesine and placebo^[2].
• The verdict that mattered most. The Reduction in Cardiovascular Events by Acadesine in Patients Undergoing CABG trial (RED-CABG) was the largest and most recent of them, run across 300 sites in 7 countries.

Bottom line: RED-CABG was stopped early, after a prespecified futility analysis, when the data made clear it had almost no chance of showing a benefit. The primary outcome, death, nonfatal stroke, or severe heart failure needing mechanical support, occurred in 5.0% of the placebo group and 5.1% of the acadesine group^[3].

That negative result is also why the trial got funded in the first place. The earlier pooled meta-analysis of the first five smaller trials had found that acadesine decreased perioperative myocardial infarction by 27% and cardiac death by 50%^[2], a signal strong enough to justify a confirmatory trial on the scale RED-CABG was designed for. RED-CABG was that confirmation, run bigger and later, and the earlier signal didn’t survive it.

AICAR didn’t fail to get studied: it got studied thoroughly, in the one indication anyone ever tested it for, looked promising on the first pass, and lost on the pass built to confirm it.

The mechanism: an AMP mimetic wearing a peptide’s reputation

AICAR’s chemistry is well mapped, and none of it involves amino acids.

What’s established: AICAR is one of the most widely used pharmacological activators of AMP-activated protein kinase (AMPK), the enzyme that functions as the cell’s energy gauge^[1]. In rat hindlimb muscle, perfusing AICAR activates AMPK, shuts down acetyl-CoA carboxylase, and drops malonyl-CoA levels. The downstream effect in that experiment was a 2.8-fold jump in fatty acid oxidation over 45 minutes, plus a measurable rise in glucose uptake^[4].

What’s overstated: a lot of popular explainers describe AICAR as simply “turning on AMPK,” full stop. A systematic review found that a growing body of work traces many of AICAR’s cellular effects, spanning metabolism, hypoxia, exercise, nucleotide synthesis and cancer biology, to pathways that don’t run through AMPK at all^[1]. The compound was the field’s default tool for studying AMPK for two decades, and the same review that documents its everyday use is the one warning other researchers not to trust it as a clean AMPK probe anymore.

That second point matters more than it sounds. If researchers who have spent two decades studying this molecule in cell culture and rodent tissue still can’t cleanly separate its AMPK effects from its non-AMPK effects, a marketing claim that it will “activate your metabolism” in a human body is not a simplification. It’s a guess dressed up as a mechanism.

The dosing nobody selling AICAR mentions

The “exercise in a pill” reputation traces to one mouse study, and the dosing in that study looks nothing like what research-chemical vendors sell.

44% increase in treadmill endurance, sedentary mice, 4 weeks of AICAR alone, no exercise PMID 18674809

Narkar and colleagues gave untrained, sedentary mice daily AICAR injections for four weeks and found that even without any treadmill training, the treated mice ran 44% longer before exhaustion than untreated controls, alongside changes in muscle metabolic gene expression^[5]. That is a real, striking result. It is also a result in mice that never set foot on a treadmill before the drug, dosed daily for a month, with no human equivalent dose ever published.

Compare that to how the drug was dosed in people:

Two dosing protocols for the same molecule

Same compound, unrelated regimens.

Human acadesine studies measured the drug in blood plasma during a single intravenous infusion built around bypass surgery, a one-time hospital dosing event rather than the sustained daily regimen used in mice^[6]. No published trial has tested AICAR dosing for exercise performance, endurance, or fat loss in human participants. The gap between the mouse protocol that built the hype and the human protocol that got tested isn’t a detail: it’s the whole story.

Side effects, and a ban built on mouse data

Here the honest answer splits into two very different evidence bases.

In the cardiac trials, the safety picture was clean. Across the pooled trials, the incidence of adverse events was similar between the acadesine and placebo groups, apart from a transient rise in serum uric acid in the drug group^[2]. Acadesine wasn’t abandoned because it hurt people. It was abandoned because RED-CABG showed it didn’t help them.

In competitive sport, AICAR carries a different weight entirely. The World Anti-Doping Agency added AICAR to its Prohibited List in 2009, on the strength of the 2008 mouse endurance data and reports of illicitly distributed AICAR circulating among athletes, not any demonstrated performance effect in trained humans^[7]. It went on the list alongside GW501516 (Cardarine to anyone who’s spent time in a supplement forum), a PPARδ agonist tested in the same mouse experiments. Both drugs earned a sport-wide ban off the back of rodent data, years before either one had a published human performance trial to its name.

No published study has monitored humans for the effects of the chronic, self-administered AICAR dosing sold by research-chemical vendors. A WADA listing is a regulatory judgment about risk and fairness rather than a clinical safety finding, and this is one case where that distinction cuts against the compound: the agency moved faster than the science did.

Where AICAR sits next to metformin

AICAR is not the only drug that works partly by nudging the AMPK pathway. Metformin does too, and the contrast is instructive.

Metformin, the most widely prescribed AMPK-adjacent drug on the planet, has decades of outcome data behind it. In the UK Prospective Diabetes Study’s 10-year post-trial follow-up, overweight type 2 diabetes patients originally assigned to metformin showed significant long-term reductions in myocardial infarction (33%) and death from any cause (27%), years after the active treatment phase had ended^[8]. That isn’t a mechanism paper or a rodent study: it’s a decade of real patients living or dying at measurably different rates.

AICAR’s evidence base runs backward by comparison. It has more completed human randomized trials behind it than almost any peptide sold online, yet every one of them tested a cardiac surgery outcome, and the biggest one came back negative. A compound can be extensively studied in humans and still have zero evidence for the claim printed on the label.

What would settle the performance question

Nobody has run the trial that would answer whether AICAR does anything for a trained human body.

A study measuring AICAR’s effect on exercise performance or body composition directly in human volunteers, the comparison its marketing implies, has never been published. Until that trial exists, every endurance or fat-loss claim attached to AICAR rests on two things that were never meant to answer that question: a sedentary-mouse dosing protocol, and a cardiac-surgery drug trial built to measure whether people survived an operation rather than whether they could run faster.

That is a thinner evidentiary foundation than real peptides that still can’t clear this bar. Most unregulated research chemicals sold for performance have the excuse of being simply untested in humans at any dose, for any purpose. AICAR doesn’t get that excuse. It was tested, extensively, in thousands of people, on exactly the wrong question, and the answer came back no.

This article is for research and informational purposes only and is not intended to diagnose, treat, cure, or prevent any disease. The peptides discussed here are sold for research use only and are not for human consumption. Nothing in this article constitutes medical advice. Consult a qualified clinician before making changes to a health, training, or supplementation protocol.