How to read a GLP-1 headline without getting fooled

A GLP-1 headline has one job: to make you click. “Ozempic linked to [scary thing].” “The surprising new benefit of Mounjaro.” Both are built in a lab — not a pharmacology lab, a headline lab — and both can sit on top of a study that says something far smaller and more careful. Here are five questions that pop the balloon in under a minute. None require a science degree.

1. Was it people, or mice?

This is the single most useful filter. A huge share of “breakthrough” stories rest on studies in mice, rats, or cells in a dish. Findings in animals are how science starts, not how it ends — most don’t translate to humans. If the article won’t tell you which it was, that silence is the answer. Say it out loud: “in mice.” Watch how much excitement survives.

2. Does “linked to” mean “caused”?

“Linked,” “associated with,” “tied to” are honest words for a real but limited kind of evidence: two things tend to occur together. They are not the same as caused. People who take a medicine often differ from people who don’t in a dozen other ways, and any of those could be doing the work. Association is a reason to look closer, not a verdict. When a headline says “linked,” mentally append: “…and we don’t yet know why.”

3. How big was it, and compared to what?

“Doubles your risk” sounds terrifying until you ask: doubles it from what? If a rare event goes from 1 in 10,000 to 2 in 10,000, the risk “doubled” and is still tiny. Relative numbers (doubled, 40% more) are designed to feel large; absolute numbers (1 extra person in 10,000) tell you what’s actually at stake. A responsible source gives you both. A scary headline gives you only the first.

“Doubles your risk” is designed to feel large. “One extra person in 10,000” tells you what’s actually at stake. The gap between those two is where most health scares live.

4. Who ran it, and who paid?

A randomized trial of thousands carries more weight than a survey, a single clinic’s chart review, or a press release about an unpublished result. And funding shapes findings: a supplement study funded by the supplement company isn’t worthless, but it earns more scrutiny. You don’t have to be cynical — just ask the question every time.

5. Is a lawsuit being treated as a fact?

“Thousands sue over [drug] injury” tells you that lawyers have filed claims. It does not tell you the drug caused the injury — that’s what the process is for. Litigation and proof are different stages, and headlines routinely blur them. Allegations are allegations until adjudicated.

Put it together

The headline move	The question that defuses it
“Breakthrough study shows…”	Was it people, or mice?
“Linked to / associated with…”	Does “linked” mean “caused”?
“Doubles your risk”	Doubled from what — what’s the absolute number?
“New study finds…”	Who ran it, and who paid?
“Thousands sue over…”	Is a lawsuit being treated as proof?

Run a striking claim through those five and most of the drama drains out. What’s left is usually a modest, real, appropriately-uncertain finding — which is what almost all genuine science looks like. The headlines that survive all five questions are the ones worth your attention.

And when you’d rather have the rung looked up for you, that’s the whole point of this site: paste the claim into the Claim Checker and we’ll show you where it actually sits, with the sources underneath. The goal isn’t to make you trust us — it’s to make you harder to fool, by anyone, including us.