How chemists test what's in GLP-1 drugs

Chemists at Hong Kong's Government Laboratory have published a single lab test that can identify and measure nine different GLP-1 receptor agonists in one 20-minute run^[1]. The method, described by Tong, Leung and Hung in the Journal of Chromatography A, covers semaglutide, tirzepatide, retatrutide, liraglutide, exenatide, mazdutide, survodutide, ecnoglutide and bofanglutide, spanning approved medicines, recently approved drugs and drugs still in Phase 3 trials. It matters less for what it tells a chemist and more for what it represents: a peer-reviewed, publicly documented way to answer the question a lot of GLP-1 buyers cannot answer for themselves, which is whether a vial actually contains what the label says.

What the new method actually measures

The technique is liquid chromatography paired with high-resolution mass spectrometry, run on a C18 column with a gradient that separates most of the nine peptides within 20 minutes. The identification step is what makes it useful for verification rather than just quantitation: the method combines a full-scan high-resolution mass measurement with fragmentation scanning, so a peptide is confirmed both by its exact mass and by the pattern it breaks into. The paper reports mass accuracy of 0.0 to 0.5 parts per million and fragment-ion coverage of 76% to 100%. That is a much tighter identity check than a simple color test or a vendor's certificate of analysis.

The quantitative side held up too. Linearity was strong (r-squared of 0.995 or better) across a 4 to 235 nanogram-per-millilitre range, with detection limits between 0.44 and 2.06 ng/mL. The authors then tested the method on real, marketed products, using a simple dilute-and-shoot preparation on liraglutide, semaglutide and tirzepatide formulations, and recovered 113.7% to 118.4% of the labelled amount^[1]. That is the kind of check regulators and manufacturers run routinely on approved products. It is not the kind of check most buyers of unregulated GLP-1 vials ever get.

Why identity testing is not a hypothetical problem

The gap this method addresses is not theoretical. In December 2023 the FDA warned that counterfeit Ozempic (semaglutide) had entered the legitimate US drug supply chain^[2]. The counterfeit units carried a genuine lot number paired with a fake serial number, and testing on the seized needles found their sterility could not be confirmed, which the FDA flagged as an infection risk. At the time of the warning, the FDA and Novo Nordisk said they were still testing the seized product and did not yet have full information about its identity, quality or safety, which is precisely the gap a fast, validated identification method like the new LC-HRMS approach is built to close.

A 2024 study published in the Journal of Medical Internet Research bought semaglutide from three online sellers that required no prescription and no medical screening of any kind, then sent the vials for independent liquid chromatography and mass spectrometry testing^[3]. Measured purity ranged from 7.7% to 14.4%, against a 99% purity claimed on every label, and every sample tested positive for bacterial endotoxin, at levels up to 8.95 endotoxin units per milligram. Two of the three sellers also shipped substantially more semaglutide-related material than the label stated, by 28.6% to 38.7% over the claimed dose. Vendors do not disclose that kind of gap voluntarily. Independent testing, of the kind the new multiplexed method makes faster and cheaper to run, is the only way anyone finds out.

What this method does not do

It is worth being precise about the limits here. The paper validated its method against marketed liraglutide, semaglutide and tirzepatide, all approved medicines with an established manufacturing chain behind them. It did not test grey-market or compounded vials, and it does not evaluate sterility, endotoxin load or manufacturing conditions, only the identity and quantity of the peptide itself. A method built for a 20-minute chromatography run also assumes a lab, an instrument and a chemist running it. It says nothing about what happens when a compounded or research-labelled vial never reaches a lab at all, which is the more common scenario for the newer, unapproved molecules in the same peptide family. Publishing a validated method also does not mean every regulator or vendor is now using it. A tool existing and a tool being applied to the products actually circulating online are two different things, and the second one lags well behind the first.

Where this leaves a reader considering a GLP-1 product

The practical read is unglamorous. Analytical chemistry like this is what regulators and manufacturers use to catch counterfeit and mislabelled product after the fact, not a tool an individual buyer can run on a vial before injecting it. The FDA's own investigators needed lab testing to establish what a seized, counterfeit Ozempic pen actually contained, and the JMIR researchers needed the same kind of testing to find that some online semaglutide vials held roughly a tenth of the purity on their label. Neither finding was visible from the packaging alone. Retatrutide remains investigational and unapproved everywhere PeptideMethods covers, which means any vial sold as retatrutide sits entirely outside the kind of quality framework this paper describes for approved semaglutide and tirzepatide products. If you are considering any GLP-1 product, the sourcing question (is this from an authorised pharmacy dispensing an approved medicine) matters as much as the dosing question, and it is one your prescriber can actually help you answer.