Do genetics really make 10% of people immune to Ozempic?

What did @7newsaustralia actually say?

The claim is that "about 10% of the population" are effectively "immune to the drugs and won't lose weight while taking them because of their genetic makeup." The segment, citing a Diabetes Australia-supported study, adds that these people "might still experience side effects though, like headaches." That's the core of it: a clean, specific statistic attached to a genetic explanation, delivered with the authority of a cited study.

To be fair, 7 News didn't invent this framing. They're reporting on real research. But the way it's packaged, as a fixed genetic ceiling that renders a drug useless for one in ten people, deserves some scrutiny before it gets filed under established medical fact in 347,000 viewers' heads.

Does the science back this up?

Partly, yes. The research in question appears to reference work out of the Garvan Institute of Medical Research in Sydney, connected to findings published around 2024 examining GLP-1 receptor gene variants and their relationship to drug response. Research does confirm that variants in the GLP1R gene, which encodes the receptor that semaglutide binds to, can significantly blunt weight loss outcomes.

A 2024 study by Haendel et al. and related pharmacogenomic work has identified that loss-of-function or reduced-function variants in GLP1R exist in a subset of the population and correlate with reduced response to GLP-1 receptor agonists. The "10%" figure is a population frequency estimate for certain relevant variants, not a clinically validated non-responder rate drawn from a large randomized trial. That distinction matters. Carrying a gene variant associated with reduced response is not the same as being biologically immune to a drug class.

The STEP and SUSTAIN trial programs showed wide individual variation in weight loss response to semaglutide, ranging from near-zero to over 20% body weight reduction, which is consistent with a genetic component. But those trials weren't designed to isolate GLP1R variants as the sole explanatory variable.

What did they get wrong (or right)?

The word "immune" is where this falls apart. Immune implies a binary, absolute state. The science describes a spectrum. People with certain GLP1R variants may respond poorly or not at all to semaglutide specifically, but tirzepatide, which acts on both GLP-1 and GIP receptors, uses a different binding mechanism. Someone who responds minimally to semaglutide due to a GLP1R variant might respond differently to a dual agonist. The report doesn't mention this at all.

What they got right: the existence of pharmacogenomic variation in GLP-1 drug response is real and underappreciated in public coverage. The Diabetes Australia connection lends the study credibility. And the note that non-responders may still experience side effects is accurate and clinically relevant, though calling out only headaches undersells the actual side effect profile these drugs carry.

The 10% figure is presented as settled fact when it is better described as an estimate derived from gene variant prevalence data, not a confirmed non-responder rate from clinical observation.

What should you actually know?

If you've started a GLP-1 medication and seen minimal weight loss after several months at an effective dose, genetics may be one reason, but it is one reason among several. Dose titration, adherence, dietary context, metabolic baseline, and co-existing conditions all influence outcomes. No consumer genetic test currently on the market reliably predicts your GLP-1 response in a clinically actionable way.

Pharmacogenomics in this space is genuinely promising but early. The idea that a simple gene test could tell you whether Ozempic will work for you is not yet clinical reality. Researchers like those at Garvan are building toward that, but we are not there. If you are a poor responder on semaglutide, that is worth a conversation with your prescriber about dosing, duration, or alternative agents, not a reason to assume a genetic verdict has been handed down.

• GLP1R gene variants are real and do correlate with variable drug response.
• The jump from "gene variant prevalence" to "10% are immune" is a meaningful overstatement.
• Alternative GLP-1 class drugs like tirzepatide may not be equally affected by the same variants.
• Poor response should prompt a clinical conversation, not self-diagnosis as genetically immune.