The Drug That Does Not Work for You

Ozempic might be the most successful drug story of the decade. Developed for Type 2 diabetes, repurposed for weight loss, now prescribed in volumes that created global supply shortages. The GLP-1 receptor agonists — Ozempic, Wegovy, Mounjaro — are close to becoming a pharmaceutical archetype: the miracle drug that works, full stop.

A paper published this week in Genome Medicine would like to add a footnote.

For roughly 1 in 10 people, these drugs don't work. Not because the dose is wrong. Not because of poor adherence. Because of genetics.

A decade-long international study led by Stanford Medicine researcher Anna Gloyn identified two genetic variants in the PAM enzyme — a protein with the full name peptidyl-glycine alpha-amidating monooxygenase — that produce what the researchers are calling GLP-1 resistance. People who carry these variants have higher circulating levels of the GLP-1 hormone than people without them. And their blood sugar responds as if the hormone isn't there at all.

More of the thing. Less of the effect. The biology reversed its expected behavior.

Why This Is Weirder Than It Sounds

GLP-1 is a gut hormone released after you eat. Its job is to signal the pancreas to release insulin, slow down stomach emptying, and reduce appetite. GLP-1 receptor agonist medications work by mimicking this hormone at concentrations far higher than the body naturally produces. The logic is straightforward: more signal, more response.

The PAM enzyme activates GLP-1 through a process called amidation — a chemical modification that increases the hormone's stability and potency. People with the PAM variants have a version of this enzyme that doesn't work properly. Gloyn's team originally expected these people to have lower GLP-1 levels, because the less-activated hormone would degrade faster.

The experiment revealed the opposite.

"What we actually saw was they had increased levels of GLP-1," Gloyn said. "This was the opposite of what we imagined we would find."

More hormone, no corresponding effect. The cells aren't listening. The researchers confirmed this in human participants, in mouse models with the PAM gene knocked out, and in analysis of clinical trial data. When they looked at actual drug trials, patients with these variants showed significantly less blood sugar reduction after six months of treatment.

The mechanism — why the cells don't respond despite elevated hormone levels — remains unknown. "That is the open question," Gloyn said. They ruled out the obvious candidates: the GLP-1 receptors are intact, the hormone can bind to them normally, the receptor's signaling pathway isn't obviously broken. The resistance emerges somewhere further downstream in the cellular machinery. The team has methodically eliminated a long list of plausible mechanisms. None of them explain it.

The Population Tool That Fails the Individual

This is a story about a particular failure mode of modern medicine: the population average that doesn't apply to you.

GLP-1 drugs were developed using population-level data — clinical trials enrolling thousands of people and measuring average outcomes. They work. The trials were unambiguous. But a population average conceals a distribution, and at one end of that distribution, roughly 10% of people carry variants that make the drug's mechanism biologically ineffective for blood sugar control.

Doctors already see this. "When I treat patients in the diabetes clinic, I see a huge variation in response to these GLP-1-based medications and it is difficult to predict this response clinically," said Mahesh Umapathysivam, an endocrinologist at Adelaide University and lead author of the study.

Until now, when a patient didn't respond to six months of GLP-1 treatment, the doctor changed the drug. The patient had no idea why it didn't work. They just tried something else.

The study suggests a path toward knowing in advance. Genetic testing for PAM variants could identify the 10% for whom GLP-1 agonists won't effectively regulate blood sugar — before six months of ineffective treatment. The question of whether these same variants affect weight loss from higher-dose GLP-1 therapy remains open; the current study focused specifically on blood sugar regulation, not obesity treatment.

---

The body is not a generic system. The drug that works for nine people in ten is not a drug that works for everyone — it's a drug that works for the person the population resembles on average. For the remaining 10%, the question isn't why the drug failed. The question is why the body is wired differently, and what else that wiring means.

Somewhere in the answer is probably another drug that does work. We just haven't found it yet.

Sources:

• One in 10 people may have resistance to GLP-1 diabetes drugs — Stanford Medicine, 2026-04-10