You're Not Broken: Why GLP-1s Don't Work for 1 in 10 People (2026 Research)

A major study published in Genome Medicine on March 29, 2026 just changed how we think about GLP-1 failure. Stanford researchers analyzed a decade of clinical data and found something specific: about 1 in 10 people carry genetic variants that reduce their body's response to drugs like Ozempic, Wegovy, Mounjaro, and Zepbound.

Not because they're doing anything wrong. Not because they need more willpower. Because of two gene variants they were born with.

If you've been on a GLP-1 for six months and the scale hasn't moved the way it was supposed to — or your appetite never quieted the way your friend described — this research matters.

What the Study Found

The Stanford team, working with large trial datasets and lab models in both mice and humans, honed in on an enzyme called PAM — peptidyl-glycine alpha-amidating monooxygenase. PAM's job is to finish-assemble several hormones your body makes, including GLP-1 itself.

People who carry specific PAM variants still produce GLP-1 in their gut. They often produce more of it, not less. But the hormone doesn't do its job properly downstream — and neither do the drugs that imitate it.

The researchers looked at blood sugar responses after six months of treatment and saw a consistent pattern: people with these variants had much smaller drops in blood glucose than people without them. The 10% figure comes from how common the variants are in the tested populations.

This isn't a rare edge case. In a room of 30 people on a GLP-1, roughly three of them are likely carrying variants that blunt their medication's effect.

What This Means for You

First, and most importantly: if your GLP-1 isn't working the way you hoped, it may not be your fault. This is the sentence we want you to sit with.

Most of us have been taught to explain slow weight loss or muted appetite changes as a personal failure. "I'm probably not sticking to the plan." "I must be sneaking calories somewhere." "Other people on Wegovy lost 20% — something must be wrong with me."

The research says: sometimes, something is genuinely different about how your body processes the drug. That's biology, not character.

Second: it means the future of GLP-1 care is going to look more like precision medicine. Dr. Julia Salem, one of the co-authors, told Scientific American that the goal is to eventually test for these variants before prescribing, so patients can skip the six-month guessing game and start on a medication likely to work for them.

We're not there yet. Genetic testing for GLP-1 response isn't standard practice, and insurance won't cover it for this indication. But the paper is one more piece of evidence that the "try it and see" model of obesity pharmacology has an end date.

How to Tell If You Might Be a Low Responder

There's no at-home test. But there are patterns clinicians watch for, and patterns you can track yourself.

Signs your GLP-1 may be working well:

• Food noise (intrusive food thoughts) dropped noticeably within 4–8 weeks
• Appetite feels muted, not just smaller
• Steady weight change of 1–2% per month after the first month
• HbA1c dropping on schedule if you have diabetes

Signs worth discussing with your prescriber:

• Six months in with less than 5% total weight change despite titration
• Blood sugar barely moving even after dose increases (diabetes context)
• Appetite and food noise feel almost identical to pre-medication
• You can describe the taste of food as vividly as you could before starting

None of these individually prove genetic resistance. Lots of things cause a slow response — stress, sleep debt, medication interactions, perimenopause, under-titration, missed doses, timing of meals relative to the injection. Your clinician will want to rule those out first.

But if you've ruled them out and you're still not seeing the response your dose should produce, it's a legitimate conversation to have.

What to Do If You Think You're a Low Responder

Do not stop your medication before talking to your prescriber. And don't assume you're in the 10% after one slow month. Six months of fair trial matters.

But here's what the research suggests you can do:

1. Track the right signals, not just the scale. The scale lags behind the real story. Track food noise intensity, appetite strength, satiety duration, and energy. If none of those are moving, that's earlier evidence than weight plateau.

2. Ask about switching drug families. There are two GLP-1 drug families in 2026: semaglutide (Ozempic, Wegovy) and tirzepatide (Mounjaro, Zepbound). Tirzepatide is a dual agonist — it also activates a separate pathway called GIP. Some people who don't respond to semaglutide respond strongly to tirzepatide. See our comparison of the two.

3. Layer behavior, don't rely on biology. This one sounds unfair when you're doing everything right and the medication still isn't pulling its weight. But behavioral work — habit stacking, environmental design, identity-based behavior change — produces real weight change independent of the drug. For low responders, it matters more, not less.

4. Make sure you're not confusing resistance with titration lag. The early doses (0.25mg, 0.5mg semaglutide) are intentionally subtherapeutic — they're there to minimize nausea while your body adjusts. If you're still at a starter dose, you haven't tested the drug yet. Our titration guide covers what each dose is meant to do.

5. Rule out the obvious first. Before assuming genetics, cover: are you injecting correctly? Is the pen stored cold? Are you spacing doses exactly 7 days apart? Is your dose timing consistent? Read our piece on why Ozempic stops working for the full diagnostic checklist.

What This Research Doesn't Mean

A few things this study is not saying, because the internet will bend it in these directions:

• It's not saying GLP-1s don't work. They work for roughly 9 out of 10 people. This is about the minority who don't respond well, not the majority who do.
• It's not saying you can predict your response from ancestry. The PAM variants occur across populations. There's no "this ethnicity doesn't respond" takeaway here.
• It's not a reason to skip the medication. Most people don't carry these variants. And for those who do, the next generation of GLP-1 drugs — including oral semaglutide (orforglipron) and triple agonists in trials — may work differently.
• It's not an excuse to not do the behavioral work. For high responders, behavior makes the results stick. For low responders, behavior is the main lever. Either way, it matters.

The Honest Framing

One of the hardest things about being on a GLP-1 is the loneliness of a slow response. You read the 20% success stories. You listen to friends describe food noise vanishing overnight. And you wait for it to happen to you, and when it doesn't, you assume the problem is you.

This research is a reminder that your body is its own story. Same drug, same dose, different biology — different response. That's not failure. That's just how pharmacology actually works.

If you're in the slow-response group, the path forward is the same one we'd recommend to anyone: honest data, clear communication with your prescriber, behavioral work that's yours to own, and a willingness to switch tools if the first one isn't right.

The research just gave you one more piece of evidence that the slow response isn't a verdict on your character. It's a data point about your chemistry. Those are very different things.

---

Frequently Asked Questions

How common is GLP-1 resistance? About 1 in 10 people carry the genetic variants that reduce GLP-1 drug effectiveness, according to the March 2026 Genome Medicine study. That's approximately 10% of the population.

Can I get tested for GLP-1 resistance? Not as standard care yet. The PAM gene variants can be detected through genomic testing, but no clinical test is approved specifically for predicting GLP-1 response. Expect this to change in the next few years as precision medicine for obesity develops.

If I'm a low responder to Ozempic, will Mounjaro work for me? Possibly. Tirzepatide (Mounjaro, Zepbound) works through two pathways instead of one, so some people who don't respond to semaglutide do respond to tirzepatide. Discuss a switch with your prescriber if you've had six months of minimal response.

Does this mean GLP-1s are overrated? No. They work for the majority of people prescribed them — producing meaningful weight change for roughly 90% of users. This research explains why the remaining 10% don't see the same results.

Should I stop my medication if I suspect I'm a low responder? No. Talk to your prescriber first. Six months is the minimum honest trial. Many things look like resistance but aren't (stress, sleep, missed doses, under-titration). Rule those out with your clinical team before drawing conclusions.

---

Matt Cole is a research co-host of the Gila podcast. This article draws from Salem et al. (2026), "Genetic predictors of GLP1 receptor agonist response," published in Genome Medicine on March 29, 2026, with additional reporting from Stanford Medicine News and Scientific American.