Is This GLP-1 Claim Peer-Reviewed? How to Check in 4 Steps

To check whether a GLP-1 claim came from peer-reviewed research, trace it back to its source: a journal article (peer-reviewed), a preprint (not yet peer-reviewed), a conference abstract (briefly reviewed, not a full paper), or a press release (written by the sponsor or institution). Then check that the paper hasn't been corrected or retracted. It takes about ten minutes and four steps.

Spoiler: "a study says" can mean any of these. They're not equally solid.

What are the four kinds of GLP-1 research sources?

Source	Who checked it?	How much detail?	Typical clue
Press release	The company or university's own team	Highlights only	"Topline results," "announced today"
Conference abstract or presentation	A conference committee, briefly	A few hundred words, a slide deck	"Presented at ADA / EASD / ObesityWeek"
Preprint	Nobody outside the authors, yet	Full paper	medRxiv, bioRxiv, "not peer-reviewed" banner
Peer-reviewed journal article	Independent experts chosen by the journal	Full methods, results, limitations, often a supplement	Journal name, DOI, "published online" date

Peer review isn't a guarantee of truth. It's a quality filter: outside experts read the methods and push back before publication. A preprint skipped that step, so far. The preprint server medRxiv says it directly: preprints "should not be relied on to guide clinical practice or health-related behavior" and "should not be reported in news media as established information."

A real example: the same trial, three stages

The oral GLP-1 orforglipron's ATTAIN-1 trial went through the classic pipeline:

1. August 7, 2025, press release. Lilly announced topline results: an average 12.4% weight loss on the 36 mg dose at 72 weeks, using the "efficacy estimand."
2. September 16, 2025, journal. The full, peer-reviewed results were published online in the New England Journal of Medicine, about six weeks later.
3. The journal's primary result: 11.2%, using the "treatment-regimen estimand," which counts everyone randomized, including those who stopped the pill.

Both numbers are legitimate. But a post that quotes "12.4%" as "the study result" is quoting the press release's preferred analysis, not the peer-reviewed primary endpoint. We explain the difference in why GLP-1 trials report dropouts.

How do you trace a GLP-1 claim to its source? (4 steps)

Step 1: Look for the ID

Real trials leave fingerprints:

• An NCT number (for example, NCT03548935 is STEP 1, NCT04184622 is SURMOUNT-1, NCT03574597 is SELECT). Paste it into ClinicalTrials.gov to see the registered design and linked publications.
• A DOI (a code starting with "10."), which points to one specific paper.
• A journal name and date.

No ID, no journal, no trial name? That's not proof it's false, but it's a reason to slow down.

Step 2: Find the paper on PubMed

Search the trial name or the drug plus the main finding on PubMed. If it's a peer-reviewed article in a biomedical journal, it will usually be there with an abstract. Preprints are mostly elsewhere (though some appear with a "preprint" label).

Step 3: Check it hasn't been corrected or retracted

On PubMed, a retracted article is labeled with the publication type "Retracted Publication," and the citation says "Retraction in:" with a link to the retraction notice. Look near the top of the record. You can also search the Retraction Watch database, which Crossref acquired and made openly available in September 2023. Together they cover tens of thousands of retractions.

Step 4: Match the claim to the paper

This is where many claims fall apart. Ask:

• Is it the same drug and dose? (Ozempic and Wegovy are both semaglutide, but they're approved for different uses at different doses.)
• Is it the same population? (SELECT studied people with existing heart disease.)
• Is it the main result or a secondary or exploratory one?
• Does the paper's own conclusion say what the post says?

Spot the trick

1. "Published research shows this supplement works like a GLP-1!"

Show answer

Check the journal and the study type. "Published" might mean a small lab or animal study, a preprint, or a low-quality journal. Even a real paper about a supplement's effect on a GLP-1 hormone level is not evidence it works like a prescription GLP-1 medicine.

2. "Brand-new study, just out today!" (link goes to a company website)

Show answer

That's a press release. It may be accurate, but it hasn't been peer-reviewed, and it's written by a party with an interest in the result. Wait for the full paper, and see our post on conflicts of interest.

3. "This 2019 paper proved it." (the PubMed record says "Retracted Publication")

Show answer

A "Retracted Publication" label means the paper was withdrawn. Retractions happen for errors as well as misconduct, but either way, the findings shouldn't be cited as evidence anymore.

Scoring: 3 out of 3 means your friends should start forwarding you things before they share them.

Does peer-reviewed mean correct?

Not automatically. Peer reviewers don't usually re-run the analysis or see the raw data. Good papers can still have narrow populations, short follow-up, or results that don't replicate. That's why the next step is reading what the paper actually found, using our five-minute headline routine and the relative vs absolute risk check.

But peer review does mean someone outside the team asked hard questions, and the methods are on the record for anyone to scrutinize. That's a lot more than a screenshot gives you.

What about claims on social media?

Most viral GLP-1 claims don't cite anything at all, or cite something that says less than the post does. Our guide to red flags in viral GLP-1 tips covers the common patterns. A good rule: if a claim would change what you do with a prescribed medicine, take the source to your prescriber or pharmacist first, not just the post.

If you save articles to discuss, a quick note in your journal (paper or an app like GLPily) with the trial name and date makes the conversation easier.

Common questions

How can I tell if a study is peer-reviewed?

Look for a journal name, a DOI and a "published" date, and find it on PubMed or the journal's site. Preprint servers like medRxiv label papers as not peer-reviewed. Press releases and conference abstracts are not full peer-reviewed papers.

Is a conference presentation peer-reviewed?

Conference abstracts are usually screened by a committee, but that's much lighter than journal peer review, and the full methods often aren't available. Many are later published as full papers.

How do I check if a study was retracted?

Search for it on PubMed and look for a "Retracted Publication" label or a "Retraction in:" link. You can also search the Retraction Watch database.

Why do press release numbers differ from the published paper?

Press releases often come out first and may highlight a different analysis, dose or time point. The peer-reviewed paper's primary endpoint is the most reliable headline number.