Cardarine (GW-501516): A Complete Research Guide

Cardarine, known in the literature as GW-501516 or GW1516, is one of the more frequently misclassified compounds in performance research. It is routinely grouped with SARMs in online discussions, but it has no affinity for the androgen receptor and no hormonal activity. It is a selective agonist of the peroxisome proliferator-activated receptor delta (PPAR-δ), a nuclear receptor that regulates fatty-acid metabolism, mitochondrial biogenesis, and skeletal-muscle fiber composition.

Mechanism: what PPAR-delta actually does

PPAR-δ is one of three PPAR isoforms (alpha, gamma, and delta/beta) that function as ligand-activated transcription factors. When a ligand binds, the receptor heterodimerizes with the retinoid X receptor and binds to PPAR response elements in the promoter regions of target genes. In skeletal muscle, PPAR-δ activation upregulates genes involved in fatty-acid uptake, beta-oxidation, and oxidative phosphorylation.

The practical consequence, in animal models, is a metabolic shift. Tissues that express PPAR-δ — muscle, adipose, and to a lesser extent liver and heart — increase their reliance on fatty acids as a fuel substrate. Narkar et al. (2008), working in the Evans lab at the Salk Institute, characterized this shift in detail and framed GW-501516 as an "exercise mimetic" because of the overlap between its transcriptional signature and the signature induced by endurance training.

It is important to separate two ideas that often get conflated. GW-501516 does not replace training. In the Narkar work, sedentary mice given the compound alone showed limited endurance improvement; the clearest effects appeared when the compound was combined with a treadmill protocol. The compound appears to amplify or extend the adaptive signal from exercise rather than produce that signal from rest.

Reported effects in rodent endurance studies

The endurance data in rodents is the most-cited part of the GW-501516 literature. In the Narkar 2008 protocol, mice receiving the compound plus treadmill training ran substantially longer and farther than training-only controls before reaching exhaustion. Follow-up work across several labs has reported similar directional findings in mice and rats, though effect sizes and protocol details vary.

Mechanistically, the rodent data line up with the transcriptional story. Treated animals show increased expression of slow-twitch (Type I) fiber markers, higher mitochondrial density in muscle biopsies, and shifts in respiratory exchange ratio consistent with increased fatty-acid oxidation. A separate line of work has reported reductions in circulating triglycerides and improvements in HDL:LDL ratios in rodent and primate models, which led GlaxoSmithKline to develop the compound as a candidate dyslipidemia therapeutic in the early 2000s.

What the rodent data does not establish:

• That the same endurance effect sizes translate to humans
• That chronic administration is safe
• That the metabolic benefits persist after discontinuation
• That subjective "feel" reports in humans reflect the mechanism rather than expectation

Human pharmacokinetic and short-term tolerability data exist from GSK's early trials, but the program was halted before long-term efficacy or safety could be established in humans. Researchers working with the compound should treat the endurance literature as hypothesis-generating rather than confirmed in the human context.

The 2007 carcinogenicity signal

Any honest reference on GW-501516 has to address the cancer finding directly, because the internet discourse tends to either dismiss it or treat it as disqualifying without examining the protocol.

In 2007, GSK terminated clinical development of GW-501516. The decision followed a two-year rodent carcinogenicity study in which animals dosed daily with the compound developed tumors across multiple organ systems — including liver, bladder, stomach, skin, thyroid, tongue, testes, and uterus. The finding was not subtle, and it was dose-dependent.

Context that matters when reading this result:

• The doses used in the two-year carcinogenicity protocol were supra-physiological by design. Chronic carcinogenicity studies deliberately use doses orders of magnitude above any therapeutic range to detect signal within the lifespan of a rodent. The lowest dose in the GSK study was reported around 3 mg/kg/day, with higher arms at multiples of that.
• PPAR agonists as a class have a complicated carcinogenicity history in rodents. PPAR-alpha agonists (fibrates) produce rodent liver tumors through a mechanism that does not translate to primates because of species differences in peroxisome proliferation response. Whether a similar species-specific argument applies to PPAR-δ and the tumor types seen in the GW-501516 study is not settled.
• Tumor development required chronic daily exposure over approximately 104 weeks — roughly the full adult lifespan of the rodent strain used.
• No human cancer cases have been causally attributed to GW-501516 exposure, but the human exposure record is small and uncontrolled, so the absence of reports is weak evidence.

The responsible reading is that the 2007 data is a real signal, not an artifact, and it is the reason no pharmaceutical company has pursued the compound since. Researchers in a preclinical context generally keep exposure windows short and doses modest specifically because of this finding.

Protocols reported in the research literature

Dosing protocols used in the research-chemical literature are shorter and lower than the GSK carcinogenicity arms, and they are built around the cancer signal rather than ignoring it.

Commonly reported ranges:

• Dose: 10–20 mg per day, taken orally, typically as a single dose. Some protocols split into 10 mg morning and 10 mg pre-training. Half-life is reported in the 16–24 hour range, so once-daily is pharmacologically reasonable.
• Cycle length: 8 weeks is the widely cited ceiling. Longer runs are not well characterized in the research context and increase cumulative exposure.
• Frequency: Daily dosing is typical because PPAR-δ transcriptional effects depend on sustained receptor activation.
• Off-cycle: Equal or longer time off between cycles. There is no evidence that "bridging" at low dose reduces risk; it simply extends exposure.

Because GW-501516 is not hormonal, it does not suppress endogenous testosterone, LH, or FSH. Post-cycle therapy protocols used for SARMs or anabolic steroids do not apply. This is one of the main reasons the compound is mischaracterized as a SARM in forum discussions — the dosing cadence and cycle length superficially resemble SARM protocols, but the underlying biology and the recovery picture are entirely different.

Stacking context reported in the literature:

• In cutting or recomposition stacks, GW-501516 is combined with selective androgen receptor modulators such as Ostarine or Andarine, where the rationale is that the PPAR-δ arm handles the metabolic shift while the SARM arm handles the lean-mass preservation.
• In endurance-focused protocols, it is sometimes paired with compounds like SR9009 (a REV-ERB agonist with a partially overlapping metabolic profile), though SR9009 has its own bioavailability questions.
• It does not require an aromatase inhibitor, a SERM, or any HPG-axis recovery compound.

Safety signals beyond the 2007 study

Apart from the carcinogenicity finding, the reported adverse-event profile in short-term research contexts is relatively sparse. Liver enzyme elevations have been reported inconsistently; the mechanism is plausible given hepatic PPAR-δ expression, but the magnitude in short protocols is usually modest. Cardiac effects in rodent models have gone in both directions — some studies report favorable changes in lipid profile and cardiac efficiency, others raise questions about long-term cardiac remodeling under chronic PPAR-δ activation.

The World Anti-Doping Agency added GW-501516 to its prohibited list in 2009 and has issued explicit athlete warnings, citing the carcinogenicity data as the basis. Athletes in tested sport should understand that detection windows for the parent compound and its metabolites extend well beyond the dosing window.

Drug-interaction data is limited. PPAR-δ cross-talk with statin pharmacodynamics, thyroid hormone signaling, and AMPK pathways is plausible on mechanistic grounds but not well characterized in controlled human studies.

How researchers approach the compound today

The working picture that emerges from the literature is consistent. GW-501516 is a well-characterized PPAR-δ agonist with reproducible rodent endurance and lipid-profile effects, a mechanistically coherent "exercise mimetic" signature, and a real but dose-dependent carcinogenicity signal from chronic supra-physiological exposure.

Researchers working with it in preclinical contexts generally:

• Keep exposure windows short (weeks, not months)
• Stay at modest doses relative to the carcinogenicity study arms
• Do not run continuous long-term protocols
• Treat it as a metabolic tool rather than a performance-enhancing drug category
• Separate the endurance literature (reproducible in rodents, uncharacterized in humans) from the lifespan-extension or body-composition claims that circulate in non-scientific sources

The compound sits in an unusual place in the research-chemical landscape: a drug with a clear mechanism, a clean transcriptional story, genuine preclinical findings, and a carcinogenicity signal that has kept it out of the pharmaceutical development pipeline for nearly two decades. All of those things are true simultaneously, and a responsible reading of the literature holds them together rather than picking the parts that support a predetermined conclusion.

Open questions

Several gaps in the literature are worth flagging for researchers designing protocols or interpreting the compound's place in the broader metabolic-agonist landscape:

• Whether the rodent carcinogenicity finding reflects a species-specific PPAR response (as with PPAR-α fibrates) or a mechanism that would translate to humans. No definitive work has resolved this.
• What short-term, low-dose exposure actually does in controlled human protocols. The GSK Phase I / Phase II data is limited, and no modern controlled trial has been run.
• Whether newer PPAR-δ agonists with different tissue distribution or shorter half-lives — seladelpar and related compounds under investigation for primary biliary cholangitis — offer a cleaner profile, and whether findings from those programs update how Cardarine should be read.
• How PPAR-δ activation interacts with concurrent androgen-receptor modulation at the transcriptional level in skeletal muscle, which is relevant to the cutting-stack protocols but not directly studied.
• Whether the endurance effect in rodents depends on the training stimulus being present, or whether there is a sedentary-protective metabolic effect worth isolating.

Research on the PPAR-δ pathway continues in adjacent indications — dyslipidemia, muscular dystrophy, and metabolic dysfunction-associated steatohepatitis — and findings from those programs may eventually clarify questions that the GW-501516 literature itself cannot resolve.