Is My GLP-1 Working? Plateaus and Averages

Somewhere around week five or six, a lot of people go quiet on the scale and loud in their own heads. The first pounds flew off, then they stopped, and now every weigh-in feels like evidence in a case you're building against yourself.

So here are the questions we see most, answered with what the trials and the FDA labels actually say. One thing to hold onto before you start: of the six GLP-1 medications approved for weight in the US, not one has a rule in its label telling you when to give up. When progress stalls, we talk through what to do next.

How much should I actually expect to lose?

Depends on the medication, and the honest answer comes with a range rather than a number.

In STEP 1, the trial behind Wegovy, 1,961 adults were randomized and the semaglutide group lost an average of 14.9% of body weight at week 68 against 2.4% on placebo (New England Journal of Medicine). In SURMOUNT-1, the trial behind Zepbound, 2,539 adults were randomized and the tirzepatide groups lost 15.0%, 19.5% and 20.9% at 5 mg, 10 mg and 15 mg over 72 weeks (New England Journal of Medicine).

The higher-dose Wegovy approved in March 2026 landed at 18.8% at week 72 in its label (Wegovy HD prescribing information). The Wegovy pill came in at 13.6% over 64 weeks, and Foundayo, the once-daily orforglipron tablet, at 11.1% at its top dose (Foundayo prescribing information).

Note that Ozempic and Mounjaro are the diabetes versions of the same two molecules, with different approved doses, so weight numbers from the obesity trials don't map onto them cleanly.

But the average isn't what happens to me, right?

Right, and this is the single most useful thing on this page. The trials published how people actually spread out, and the spread is wide.

In STEP 1, at week 68: 83.5% lost at least 5%, 66.1% lost at least 10%, 47.9% lost at least 15%, and 30.2% lost at least 20%. On placebo those figures were 31.1%, 12.0%, 4.8% and 1.7%.

In SURMOUNT-1 at the 15 mg dose: 90.9% lost at least 5%, 83.5% at least 10%, 70.6% at least 15%, 56.7% at least 20%, and 36.2% at least 25%.

Notice what those two lists mean in both directions. Roughly one in six people on semaglutide did not reach 5%. And more than a third of people on high-dose tirzepatide lost a quarter of their body weight. Both of those people were in the same trial reading the same average.

A real-world study of 483 people on semaglutide or liraglutide found 17.8% had lost under 5% at an average of 17 months (BMJ Open). So the trial spread holds up outside the trial.

Why does the ad say a bigger number than my pharmacist's printout?

Because trials get analyzed more than one way, and marketing tends to pick the flattering version.

Without going too far into the weeds: one analysis counts everyone who was randomized, including people who quit the drug or dropped out, and one counts what happened in people who actually stayed on it. The first is more conservative and is generally what ends up on the FDA label. The second produces a bigger number and often ends up in a press release.

A live example. When high-dose Wegovy was approved, the manufacturer's announcement led with 20.7% mean weight loss. The FDA label for the same trial says 18.8%. Neither is a lie. They're different analyses of the same 1,407 people. If you want to know what to expect, the label number is the safer one to plan around.

I've stalled at week six. Is something wrong?

Almost certainly not, and week six is early enough that you may not be on a therapeutic dose yet.

Look at the ladder. Wegovy starts at 0.25 mg for four weeks, then 0.5 mg for four weeks, then 1 mg, then 1.7 mg, and reaches the 2.4 mg maintenance dose at week 17 (Wegovy prescribing information). Zepbound starts at 2.5 mg, which its label says outright is "for treatment initiation and is not approved as a maintenance dosage" (Zepbound prescribing information).

At week six you're on the second rung of a five-rung ladder. The trial results you read about came from people who had been at full dose for a year or more.

Weight loss also isn't linear for anyone. It comes in steps, with flat stretches in between, and the flat stretches feel much longer than they are.

How do I tell a real stall from normal scale noise?

By giving it more time than feels comfortable, because the noise is bigger than most people realize.

Researchers measured two-week weight change in 46 adults who were in energy balance, meaning not gaining and not losing. The average change was 0.26 kg, with a standard deviation of about 1.2 kg (Physiological Reports). In plain terms, a swing of two or three pounds over two weeks is what happens in people whose weight is genuinely stable.

There's also a weekly rhythm. A study of 80 adults across 4,657 daily weigh-ins found weight peaks on Sunday and Monday, falls through the week, and climbs again from Saturday. The authors put it plainly: "Weight variations between weekends and weekdays should be considered as normal instead of signs of weight gain" (Obesity Facts).

Practical version. If you weigh yourself Monday morning after a weekend, you're measuring the top of your own cycle. Compare weekly averages against weekly averages, or compare the same weekday to itself, and give any apparent stall three to four weeks before you call it one.

It's week twelve and I haven't lost 5%. Should I stop?

This is the question we most want to answer carefully, because the data says something encouraging and a lot of internet advice says the opposite.

Researchers went back into SURMOUNT-1 and found 18% of participants had not reached 5% weight loss by week 12. Of those so-called late responders, 70% got to at least 5% by week 24, and 90% got there by week 72 (Diabetes, Obesity and Metabolism).

Nine out of ten people who looked like failures at week 12 were not failures. If a week-12 checkpoint had been used to pull them off treatment, most would have lost something that was still coming.

There's a similar signal on semaglutide, though weaker. A post hoc look at STEP 4 using a 5% threshold at week 20 found that the test was wrong about who would fail more often than it was right, and the people it flagged still lost about 6% by week 68 anyway. We're describing that one loosely on purpose, because it was presented as a conference abstract rather than a full paper.

None of this means you should ignore a genuine lack of response for a year. It means week 12 is too early to conclude anything, and that conversation belongs with your prescriber rather than with a forum.

Is there an official rule for when a GLP-1 isn't working?

For one older medication, yes. For the ones most people are taking, no. This surprises people, so we will be exact about it.

Saxenda, the daily liraglutide injection, has this instruction in its label: evaluate weight change 16 weeks after starting, and discontinue if the patient hasn't lost at least 4% of baseline body weight, "since it is unlikely that the patient will achieve and sustain clinically meaningful weight loss with continued treatment" (Saxenda prescribing information).

Now the part that matters for most readers. Wegovy, high-dose Wegovy, the Wegovy pill, Zepbound, Ozempic, Mounjaro and Foundayo contain no equivalent instruction. No percentage, no week, no stopping rule. The only response-related language in those labels is a general note to consider treatment response and tolerability when picking a maintenance dose.

So where does the widely repeated "5% by 12 weeks" test come from? Europe, and a different drug. The European product information for Saxenda says to stop after 12 weeks on the 3 mg dose if the patient hasn't lost at least 5%, and the European Wegovy document has a 12-week rule for adolescents only (European Medicines Agency). It's a real rule. It just isn't a US semaglutide or tirzepatide rule, and it gets quoted as though it were.

There is also no agreed clinical definition of a non-responder at all. A 2025 review notes the definition "varies slightly between clinical studies" (Frontiers in Endocrinology). The 5% and 4% thresholds you'll see come from older guidelines written before these medications existed.

Should I ask to move up a dose?

Maybe, and the labels are specific about the timing, which is useful to know before you ask.

Wegovy injection moves in four-week steps on a fixed schedule. Its label also says that if you don't tolerate a step, consider delaying escalation by four weeks. The 7.2 mg dose is not a routine rung; the label says it's an option after tolerating 2.4 mg for at least four weeks, which puts the earliest arrival around week 21.
Zepbound goes up in 2.5 mg increments after at least four weeks at the current dose, to a maximum of 15 mg. Only 5, 10 and 15 mg are approved as maintenance doses for weight.
The Wegovy pill moves every 30 days, from 1.5 to 4 to 9 to 25 mg.
Foundayo moves every 30 days, from 0.8 mg up to a maximum of 17.2 mg once daily. If you miss seven or more doses in a row, its label says to restart escalation at a lower dose.

Two things to raise with your prescriber rather than decide alone. Going up faster than the label schedule tends to buy nausea rather than results, and the highest dose is not automatically the right dose. In SURMOUNT-1, the 10 mg group lost 19.5% and the 15 mg group lost 20.9%, which is a real difference and a smaller one than the dose jump suggests.

If you're thinking about changing how you take your dose rather than how much, our piece on splitting a weekly dose into two covers what the pharmacology does and doesn't support.

When does it flatten out for good?

Later than week six, and the curve does eventually level.

Two trials make the point cleanly. STEP 1 ran 68 weeks and ended at 14.9%. STEP 5 ran 104 weeks in 304 people and ended at 15.2% (Nature Medicine). Between week 68 and week 104, the average barely moved. The steep part of the work happens across roughly the first year to fourteen months.

We're deliberately not giving you a specific plateau week, because none of these papers states one in the text. It shows up as a curve flattening in a figure, and any precise week you see quoted is somebody's estimate from a graph.

As for why it flattens, part of it is straightforward physics. A smaller body burns fewer calories. There's also a measured metabolic adaptation on top of that: in a classic study of 41 people, maintaining a weight at least 10% below starting weight reduced total energy expenditure by roughly 6 to 8 kcal per kilogram of lean mass per day beyond what the smaller size alone would predict (New England Journal of Medicine). Your metabolism isn't broken. It's smaller, and it's running slightly more efficiently than it used to.

Does the scale even matter most?

It's the easiest thing to measure, which isn't the same as the most useful thing to measure.

A review of 40 studies that used DXA body scans found that on average, 29.1% of the weight lost on these medications is fat-free mass rather than fat (Obesity). The authors describe that as the upper range of what non-surgical weight loss produces. Fat-free mass includes muscle, and muscle is the part you'd rather keep.

Two things move that number in your favor, and neither requires a prescription. Getting enough protein, which is harder than it sounds when your appetite is gone, and doing some resistance work. Our pieces on the federal protein target and on why people move less on a GLP-1 both go deeper.

Practically, if the scale hasn't moved but your waistband has, your stairs feel easier, or your labs improved, something is working that the scale can't see.

What happens if I just stay on it?

The best data on that arrived in 2026 and it's genuinely informative.

SURMOUNT-MAINTAIN put 441 people on tirzepatide for 60 weeks, then randomized 378 of them to stay at their maximum tolerated dose, drop to 5 mg, or move to placebo, and followed them to week 112 (The Lancet).

At week 112, measured from the study's start: the maximum tolerated dose group was down 21.9%, the 5 mg group down 16.6%, and the placebo group down 9.9%. The clearest number is rescue therapy, meaning people who regained at least half of what they'd lost. That happened to 8% on the full dose, 25% on the reduced dose, and 67% on placebo.

Two takeaways. Staying on treatment holds the result, and a reduced dose holds a meaningful part of it. Stopping outright is where most of the regain happened.

Final Takeaway

If you're at week six and discouraged, the numbers are mostly on your side. You're probably not at a full dose yet, a two to three pound swing over two weeks happens in people whose weight is perfectly stable, and 90% of the tirzepatide patients who looked like non-responders at week 12 got to 5% or better by week 72.

Compare weekly averages rather than individual mornings, give a suspected stall a month before you name it, and remember that no US label for these medications contains a rule about when to quit. If yours truly isn't working, that's a conversation with your prescriber about dose, timing and expectations, not a verdict you hand down to yourself on a Monday morning.

If you enjoy posts like these, you can subscribe to receive newsletter updates.

Frequently Asked Questions

How often should I weigh myself?

Whatever cadence keeps you calm and consistent. If daily weighing sends you into a spiral, weekly on the same weekday is plenty. If you do weigh daily, use the weekly average and ignore the individual days, because the day-to-day movement is mostly water, sodium and what your digestive system happens to be holding.

Does where I am in my menstrual cycle affect the number?

Fluid retention does shift across the cycle, and it's a well documented experience. We're not attaching a pound figure to it, because the study most often cited for that measured self-reported fluid retention on a rating scale rather than actual body weight. Compare like to like across cycles rather than across weeks.

I lost a lot in month one and almost nothing in month two. Did I break something?

No. Early loss usually includes a fair amount of water and reduced digestive contents, so the first few weeks tend to overstate the trend and the following weeks understate it. The trials show the loss continuing for a year or more.

Is a plateau a sign I need a higher dose?

Sometimes, and not always. The labels put four weeks between injectable steps and 30 days between oral steps for a reason, and the top dose isn't automatically better for any individual. Bring your actual weight log to your prescriber rather than a single frustrating weigh-in.

If I've lost less than 5% at four months, is it over?

Not by any US label standard, because none of these medications has one. Saxenda is the only exception, at 4% by 16 weeks. For everything else this is a clinical judgment call, and it should account for your dose, how long you've been at it, side effects, and what else has changed in your health.

Is My GLP-1 Actually Working? Answers to the Questions New Users Ask at Week Six