If You Don’t Lose Weight on GLP-1s, You’re Not an Idiot. Read This.

I have a patient who lost 40 pounds on the first two doses of a GLP-1. She came back for her follow-up wearing clothes she had bought that week, half-apologizing, as if she had cheated somehow.

I have another patient who did everything right. Weekly injection, on time, branded product, rotating sites, protein at every meal, on the highest dose for months. She lost seven pounds. She sat in my office and said, quietly, “I think I’m the one person this doesn’t work on. I think I’m just broken.”

She is not broken. She is not lazy. She is not an idiot. And she did not “fail Ozempic.”

I hear some version of that sentence every week now. Enough that I want to write down, once, in one place, what I tell people in the exam room. Because the gap between the 40-pound patient and the 7-pound patient is not willpower. It is biology, pharmacology, dose, time, and a handful of fixable mistakes that nobody explained to them.

Read this before you quit the drug or start blaming yourself.

First, the honest numbers

In the large clinical trials, roughly 10 to 15 percent of people lose less than 5 percent of their body weight. That is the rough definition of a true non-responder.

In the real world the number looks worse, closer to one in five. But a large share of those “non-responders” never got a fair trial. They were still on a starter dose. They stopped because of nausea before they ever reached a therapeutic dose. They used compounded product of uncertain potency. They were still drinking their calories. They were on a second medication quietly working against the first.

So hold two things at once. True biological non-response is real. It is also not the majority of stalled cases. Most of what I see is a fixable problem wearing the costume of a failed drug.

What the drug actually does (and does not do)

GLP-1 medications do not melt fat. They do not speed up your metabolism. They are keys.

The lock is the GLP-1 receptor, a G-protein-coupled receptor sitting on the surface of cells in a few specific places:

• The brainstem and hypothalamus, where appetite, “food noise,” and reward are processed

• The vagus nerve, which carries the gut-to-brain fullness signal

• The stomach and intestine, where it slows emptying

• The pancreas, where it raises insulin and lowers glucagon

Your own GLP-1 is a meal hormone made by cells in the gut. It lasts minutes. The drugs last a week because they were engineered so the body cannot chew them up quickly.

Weight loss on these drugs is mostly brain-driven reduced intake. The molecule reaches receptors in regions of the brain that sit outside the blood-brain barrier and turns down wanting, portion size, and the background chatter about food.

That is the mechanism. Now notice what it implies. If your weight problem is driven mostly by that gut-to-brain satiety circuit, the key fits the lock and the door opens. If your weight problem is driven by something else, the key still turns. It just was not the main door.

The receptor is not a static lock

This is the part almost nobody explains, and it changes how you should think about switching drugs.

After the drug binds, the GLP-1 receptor can do several things:

1. Signal through a pathway that produces cAMP. This is the useful “I’m full, release insulin” signal.

2. Recruit a protein called beta-arrestin.

3. Get pulled inside the cell, a process called internalization.

4. Either recycle back to the surface or get degraded.

Drugs that favor the cAMP signal and keep more receptor on the cell surface tend to keep working. Drugs that pull the receptor inside faster can desensitize the cell over time.

This is why two “GLP-1 drugs” are not interchangeable.

Semaglutide (Ozempic, Wegovy) is a potent GLP-1 receptor agonist. A strong key for one lock.

Tirzepatide (Mounjaro, Zepbound) is a dual GIP/GLP-1 agonist. At the GLP-1 receptor it behaves as a biased agonist: strong cAMP signaling, relatively weak beta-arrestin recruitment, less internalization. On top of that it activates GIP receptors in fat tissue and in brain circuits semaglutide does not touch.

That is one reason a meaningful number of people who stall on semaglutide finally move on tirzepatide, and why a smaller group does the reverse. When you switch, you are not switching brands. You are changing which receptors fire and how long they stay available on the surface of the cell.

Genetics change the lock

Your GLP-1 receptor is built from a gene, and that gene varies.

Certain variants in the GLP-1 receptor gene are associated with more weight loss on these drugs, likely because more receptor gets trafficked to the cell surface. The same variants also track with more nausea. Better lock, louder side effects.

Other variants do the opposite: fewer receptors on the membrane, weaker binding, or loss of function along a specific signaling pathway. In the laboratory, very high drug concentrations can sometimes rescue a weak receptor. In the clinic, that looks like the patient who needs the top dose and still loses only a little.

Variants in the GIP receptor gene matter more for tirzepatide, including who gets the worst GI side effects.

There is a second genetic story that is not about the receptor at all. Roughly one in ten people carry variants in a gene called PAM that create something like GLP-1 resistance: they run higher levels of their own GLP-1, but the hormone does less work. More key in the blood, weaker effect at the lock.

Two cautions. The weight-loss data on these variants is still thinner than the glucose data. And across the whole population, genetics explain less of the variation between people than dose, sex, diabetes status, and behavior do. So do not treat a consumer DNA screenshot as your destiny. Treat it as one more reason two humans are not the same experiment.

Obesity is not one disease

Layer phenotype on top of the receptor. This is the most useful clinical map I use, adapted from the obesity research group at Mayo Clinic.

Hungry gut. You eat a normal portion, feel full, and are hungry again 90 minutes later. Fast gastric emptying, often low levels of your own GLP-1 and PYY. These are the best responders to GLP-1 and tirzepatide. In one study, a hungry-gut subtype lost roughly 21 percent of body weight at six months on tirzepatide versus about 12 percent in the other groups. You are replacing a hormone the gut under-makes.

Hungry brain. You never feel full. Second and third plates. The receptor drug helps some, but brain-acting combinations such as phentermine-topiramate often outperform a GLP-1 alone.

Emotional hunger. Food as mood regulation, reward, stress relief. GLP-1s quiet the noise for many people. They do not treat the feeling that started the eating.

Slow burn. Lower resting energy expenditure. The drug cuts intake. If you also lose muscle along the way, your metabolism falls further and the scale dies.

Many patients are a mix. That is why “I did everything and lost four pounds” and “I blinked and lost forty” can both be true stories about the same drug.

Before you call yourself a non-responder

Most “failures” I see are still one of these:

• Still on a starter dose. 0.25 mg of semaglutide and 2.5 mg of tirzepatide exist so your stomach survives week one. They are not the working dose.

• Not 12 weeks at a therapeutic dose. For semaglutide that usually means 1.7 to 2.4 mg. For tirzepatide it is often 10 to 15 mg. Time spent at a low dose does not count as a fair trial.

• Missed or late injections. Drug levels fall. Inconsistent levels look like “it stopped working.”

• Compounded product with uneven potency, or a pen left in a hot car.

• The same injection site every week until the tissue is lumpy and absorption drops.

• Liquid calories. Lattes, juice, alcohol, “healthy” oils, grazing. Appetite fell. Intake did not fall enough.

• Protein so low you are losing muscle. Smaller engine, slower loss, worse plateau.

• Interfering conditions and medications. Sleep apnea, low thyroid, steroids, insulin, many antidepressants, antipsychotics, some blood pressure drugs.

• Type 2 diabetes. Average loss is consistently smaller than in people using the drug for obesity alone.

• Sex. Men, on average, lose less than women on the same dose.

Higher body weight can also mean lower circulating drug levels at the same milligram dose. The person who most needs the drug to work may be the one getting the least exposure.

The protocol I want you to screenshot

Work through this in order, with your prescriber.

1. Confirm the basics. Branded or quality-assured product. Weekly, on time. Rotate sites. Twelve full weeks at the dose that is supposed to work, not twelve weeks total.

2. Protein first. Target roughly 1.2 to 1.6 grams per kilogram of body weight per day. For many adults that is 80 to 120 grams or more. Eat it at the start of the meal. Use a shake if solid food volume makes you sick. This is the single highest-yield nutrition change on these drugs.

3. Lift two to three times a week. Walking helps your heart. Resistance training tells your body to keep muscle so your metabolism does not collapse under you. Professional guidance now treats this as part of GLP-1 care, not an optional extra.

4. Seven honest days of logging. Food and drink. If you are not logging liquids, you are not logging.

5. Sleep seven hours. Treat apnea. Stress and five hours a night will fight the receptor every day.

6. Bring every medication and get basic labs. TSH, A1c, and the rest of the metabolic picture. Do not guess which pill is blocking you.

7. Still under 5 percent after a fair max-dose trial? Switch class. Semaglutide and tirzepatide are not the same key. A meaningful group who stall on one move on the other.

8. Still stuck? Add a second mechanism for the phenotype you actually have. Hungry brain often needs a brain-acting add-on such as phentermine-topiramate. Emotional eating needs the behavior treated, not a higher milligram. Naltrexone-bupropion is used in some obesity clinics for this reason.

9. High-risk BMI that will not move after optimized medication? Surgery is another tool. It is not a moral verdict.

10. Remember that some people are slow responders, not non-responders. A portion of early “failures” reach 5 percent or more at later time points if they stay on treatment.

What “working” actually looks like

So you stop measuring yourself against the 40-pound outlier:

• Food noise drops before the scale does. That is the receptor in the brain.

• Portions shrink without a speech about discipline. That is satiation.

• The gap between meals stretches. That is satiety and slower emptying.

• Clothes change before the number does. Muscle and water hide fat loss.

• Five percent at a real therapeutic dose is a clinical response. Blood pressure, glucose, reflux, and sleep often improve even when social media would call it “nothing.”

• A stall after early loss is usually a smaller body defending its new weight, not a dead pen. Recalculate intake and protect the muscle.

Here is the red flag that is not a normal plateau: hunger and food noise never changed at all on a high dose. That points to exposure, absorption, phenotype, interfering medications, or a lock the current key does not fit. It does not point to your character.

Save this

The drug is a key. The GLP-1 receptor is the lock. Your phenotype is which door is jammed. Dose, protein, muscle, sleep, and other medications decide whether the key even reaches the door.

Forty pounds on two pens and four pounds on the highest dose can both be honest biology.

If you are the four-pound person, you are not broken and you did not fail the medication. You need a different key, a second lock, or a different door.

The drug changed the field. It did not make obesity simple. And you are allowed to need a different plan.

Send this to the person who thinks they “failed” their GLP-1. They probably did not.

Blessings.

Afshine Ash Emrani, M.D., F.A.C.C.
Assistant Clinical Professor, UCLA
David Geffen School of Medicine

Join Our new website

Castle-Connolly Nationwide Top Doctor (Since 2008)
Los Angeles Magazine Super Doctor (Since 2010)
LA Style Magazine Top 100 Doctors in America (2024)

Subscribe to mynewsletter: substack.com/@afshineemrani
Explore my books: Amazon Author Profile