Even the Muffler Wears Out: The GLP-1 Tolerance Problem Nobody Wants to Talk About

“Should I call you Mrs. Guinea or Mrs. Pig?”

Sayer Ji just published another fantastic takedown of the GLP-1 craze. If you haven’t read “Silencing the Alarm,” do it today…

He asked the question almost nobody in mainstream medicine will: why does one in eight American adults now need a weekly injection for a condition that barely existed at population scale two generations ago?

He walked through the withdrawal trials showing the weight comes right back when the drug stops. He showed that roughly 40% of the weight lost is lean mass. He separated established harms from contested ones, which is more intellectual honesty than you’ll ever get from the drug’s marketing department. And he lands the metaphor perfectly: these drugs are a mute button on a smoke alarm while the house keeps burning — similar to metaphors I’ve been using with patients for decades.

But one number in his piece caught my eye, and I don’t think he followed it all the way to where it leads…

Ji pointed out that your body’s own GLP-1 survives about two minutes in the blood before it’s broken down. While artificial Semaglutide (Ozempic, Wegovy) lasts about a week. Divide one by the other, and you get his figure: the drug’s signal lasts roughly 5,000 times longer than the body’s own.

He used this number to show these drugs don’t “mimic” a natural hormone; they replace a brief conversation with a continuous, deafening tone. True. But here’s the question that kept nagging me: what does the body do with a tone that never stops?

You don’t have to be a rocket scientist or a medical doctor to know the answer. It turns down the volume…

You Already Know How This Works

Most of you have done this, or watched someone else do it.

You get a head cold, you buy a bottle of nasal decongestant, and it’s a miracle. But three days later it works a little less. A week later you’re spraying constantly — more stuffed up than when you started.

That’s not your imagination, and it has a name: rebound congestion, or rhinitis medicamentosa. The UK’s drug regulator issued a formal safety warning about it, and specifically labeled it under tachyphylaxis, the medical term for a drug that stops working the more you use it.

SIDE NOTE: This is not a new problem…

Doctors described the rebound effect of decongestant sprays in a 1931 study, and by 1945 JAMA was reporting dozens of patients whose ongoing congestion was being induced by the very spray they used to treat it. A year later, Mayo Clinic named the condition rhinitis medicamentosa.

I only know about this because my late mother told me about getting hooked on nasal spray for a short time back in the mid-1960s, shortly before I was born. I say “short time” because she realized it and took herself completely off them right away.

The spray works by hammering a receptor in your nasal lining…

Hammer any receptor long enough, and the cells defend themselves: they pull receptors off the surface, dampen the signal inside the cell, or both. Same story with caffeine. Same story with sleeping pills. Same story with opioids. Same story with insulin. Same story with (insert any number of medications here____________).

It’s one of the most basic rules in physiology: a signal that never shuts off stops being a signal.

The GLP-1 receptor is no magical exception. It belongs to the same big family of receptors (G protein-coupled receptors) that are notorious for this very behavior. So the question isn’t whether it adapts. The question is how fast, and where.

Here are the studies that answer this question…

Tolerance by Lunchtime

In 2011, Michael Nauck, one of the most prominent incretin researchers in the world (no fringe outsider), put nine healthy volunteers on a continuous IV drip of GLP-1 for eight and a half hours and fed them two meals.

One of GLP-1’s main jobs is slowing how fast your stomach empties. That’s a big part of why these drugs make people feel full, and why they cause the nausea, bloating and, in the worst cases, the gastroparesis (dramatically slowed intestinal motility) that Ji discussed in his article.

After the first meal, the brake on the stomach worked. After the second meal, same drip, same day, it was significantly weaker. Blood sugar, which dropped after the first meal, actually rose after the second. The authors’ conclusion: the stomach-slowing effect is subject to rapid tachyphylaxis, and the drug’s control of after-meal blood sugar weakens with continuous use.

A follow-up from the University of Adelaide, with Nauck as coauthor, nailed it down…

They compared a continuous 24-hour infusion against two separate infusions with a break between them. The on-and-off pattern, the one your body actually uses, kept working. The continuous one had faded within about 20 hours. The researchers concluded that short-acting versions of these drugs may actually beat the long-acting ones for controlling after-meal blood sugar.

Read that again. The very design that makes a once-a-week shot possible, the very thing behind Ji’s 5,000X figure, works against at least one of the drug’s own effects.

The Question I Kept Asking

I’ll be transparent about how this article came about. After reading Ji’s piece, I sat down with Claude and started pushing…

My first question was the obvious one for anyone who has watched insulin resistance develop in patient after patient for 35 years — increasingly so due to the nature of the SAD (Standard American Diet). If you flood a receptor around the clock, where’s the point at which it becomes saturated? In other words, where does a person stop being GLP-1 sensitive and start becoming GLP-1 resistant?

Because physiologically, it has to happen.

Saturation and resistance aren’t the same thing, but they are related (and they are also related to the need for more medicine — see my mom’s temporary addiction). And the numbers turned out to be more interesting than I expected…

First, a small correction that actually strengthens Ji’s case. His 5,000X is a measure of how long the signal lasts, not how loud it is. That matters, because the drug never gets as “loud” as you might think.

Semaglutide rides around in your blood stuck to a carrier protein called albumin. More than 99% of it, according to the FDA’s own label. Only the small unbound fraction (less than 1% for those keeping score at home) is free to bind to the receptor. Run the numbers on how much free drug is circulating at the standard 2.4 mg dose against how tightly semaglutide grabs the receptor, and the receptors aren’t close to full. Roughly half to two-thirds are occupied at any given moment. Around the clock. For months and years.

If that sounds like good news, it isn’t…

Think of a parking lot. Your own GLP-1 is a rush of cars after a meal that fills spaces for a couple of minutes, then empties out. The drug is a fleet that keeps the lot half to two-thirds full, 24 hours a day, and never leaves. What prompts a cell to start hauling receptors in off the surface isn’t just how full the lot gets for a moment. It’s how long the cars stay parked in the lot.

So the drug lands in the worst of both worlds: never enough occupancy to max out the effect, but constant enough to wear the system down. That’s the recipe for the parking lot itself getting smaller. And that’s where saturation and resistance meet — the answer to the question I was asking Claude about.

The Drugmakers Already Know This

This next part requires no pharmacology at all.

It’s another of those dirty little secrets that I didn’t bother to address in my takedown of GLP-1’s a few months ago, describing a phenomenon I had two different patients discuss with me on Friday…

In 2018, researchers at Imperial College London reported in Nature Communications that experimental GLP-1 drugs designed to avoid being pulled into the cell produced more sustained insulin release than the FDA-approved GLP-1 drugs on the market at that time.

Real-world Translation: Receptor-shutdown limits how well the approved drugs work…

Two years later, Eli Lilly’s own scientists published that tirzepatide, sold as Mounjaro and Zepbound, the drug that just out-earned Keytruda, is “biased” at the GLP-1 receptor. It favors the signal they want while recruiting less of the cellular machinery that drags receptors in off the surface. Their own experiments with pancreatic islet cells showed that this shutoff machinery limits the insulin response to native GLP-1, but not to tirzepatide.

Nobody engineers a workaround for a problem that doesn’t exist. Lilly built one because receptor tolerance is real, it’s measurable, and it costs them efficacy. That isn’t my speculation. It’s their paper.

How well did their workaround actually work?

Three Times the Drug, Three More Points

If the receptors were already doing everything they could, pushing in more drug would accomplish nothing. That’s not what’s happening. Novo Nordisk’s STEP UP trial tripled the semaglutide dose, from 2.4 mg to 7.2 mg weekly, and got average weight loss of 20.7% instead of 17.5% at 72 weeks. Tripling the drug purchased roughly three percentage points.

Not surprisingly, Novo called that a win…

Look at it the other way: three times the dose of a drug whose established harms include pancreatitis, gallbladder disease, bowel obstruction and gastroparesis, and more, were traded for a few more points on the scale.

Meanwhile, weight loss in the big trials flattens out after roughly a year to a year and a half. That’s usually explained as the body (cough cough) “finding a new set point.” Maybe. But there’s a second explanation sitting right in front of us that nobody is testing: a receptor system adapting to a signal that never shuts off.

And the industry’s pipeline tells you which way they’re betting…

Higher doses. Dual agonists. Triple agonists that hit three receptors at once. When your first target starts losing ground, you add targets. It’s the nasal decongestant playbook: switch brands and spray more often.

The Sickest Patients Start Already Behind the Eight-Ball

One more piece. The people these drugs were first approved for, type 2 diabetics, start with fewer working receptors to begin with…

Two-decade-old research in rats showed that chronically high blood sugar itself dials down GLP-1 receptor expression in the insulin-producing cells of the pancreas, and that receptor levels recovered once blood sugar was brought back down. And a 2015 study traced how high glucose strips the receptor off the cell surface, and noted that GLP-1 drugs show a weaker dose response in type 2 diabetics than in healthy people. In other words, it’s not like this is a new finding.

Notice the irony…

The metabolic environment Ji described — the ultra-processed diet and chronically high blood sugar — degrades the very receptor the drug depends on in the first place. And what restored those receptors in the animal studies? Getting blood sugar down. That’s Ji’s root-cause argument, coming at it from the receptor’s side. And his point is spot-on. The drugs are doing that in an artificial manner, without addressing the underlying metabolic dysfunction.

Turn that durned fire alarm off — it’s driving me crazy!

The Question Nobody Is Asking

Here’s what’s established…

GLP-1 tolerance happens. It can happen fast. The researchers who study these hormones have documented it, and the companies that make the drugs have engineered their newest molecules specifically to get around it to try to get around it.

To be fair about the limits: the strongest evidence for fast tolerance is for the stomach-slowing effect. The appetite effect in the brain appears more durable, which is why the drugs keep working for many people well past a year. But “more durable” is not “immune.” And “we haven’t studied that yet” is not “it’s fine.”

Here’s what nobody knows, because nobody has studied it…

What happens to your own GLP-1 system after five or ten years of being chemically overridden? Your gut still makes GLP-1. The nerve fibers that carry the “I’m full” message from gut to brain still exist. Do they work the same after a decade of being drowned out? Saturated? Who knows?

“Should I call you Mrs. Guinea, or Mrs. Pig?”

When someone finally stops the drug — and more than half of real-world users quit within a year — does their own fullness signal come back at full strength? The withdrawal trials Ji cites show the weight comes back. They don’t tell us whether the alarm itself was damaged.

Ji is right that these drugs silence the alarm while the house keeps burning. I’d add one thing…

Muffle any alarm long enough, and the body stops listening to it. And the companies selling the muffler couldn’t care less whether that alarm ever works again or not. Why would they? A patient who can’t hear the alarm is a customer for life.