Nothing. No appetite change, no food noise going quiet, no early fullness — not a diminished effect, an absent one. You have injected several times and the compound may as well be saline.
The conclusion people reach fastest is the one they should reach last: I must be a non-responder. True biological non-response exists. It is also, by a comfortable margin, the least likely explanation for what you are describing, and jumping to it means abandoning four more probable causes without checking any of them.
Work the list in order. It is ordered by how often each turns out to be the answer, and the first item alone accounts for most cases.
1. You are underdosed, or simply early
This is the answer most of the time. It is unglamorous and people hate it, and it is still the answer most of the time.
The starting doses are not therapeutic doses. Semaglutide starts at 0.25mg weekly and the obesity trials ran maintenance at 1.7-2.4mg — the starting dose is roughly a tenth of where the effect lives. Tirzepatide starts at 2.5mg against studied maintenance of 5, 10 and 15mg. Those low steps exist to let your gut adapt, not to produce weight loss. Feeling nothing at 0.25mg semaglutide is not a failure signal. It is the expected observation.
The second half of this is timeframe. The trials that produced the famous percentages ran 72 weeks. Not eight. The curves in those papers are still descending at week 40. Someone three weeks in, on step one, concluding the compound does not work for them has not run the experiment — they have read the first data point and called the study.
- 1
Weeks 1-4: starting dose
A fraction of maintenance. Many participants report little or no appetite change here. This is normal and is not predictive of anything.
- 2
Weeks 4-16: escalation
Stepping up every four weeks. Effect typically becomes noticeable somewhere in this window, and where varies enormously between individuals.
- 3
Weeks 16-36: approaching maintenance
At or near the studied maintenance range. If nothing at all is happening by here, at a proper dose, the question becomes legitimate.
- 4
Weeks 36-72: where the headline numbers were made
The trial curves keep descending. The published percentages are endpoints of a year-plus, not of a quarter.
There is also accumulation. With a roughly one-week half-life dosed weekly, the compound builds for several weeks before the level plateaus — so even at a fixed dose, week five is a higher exposure than week two. Changing nothing and waiting is a real intervention, and it is the one most likely to resolve this.
How to tell if this is you: are you below the studied maintenance range, or fewer than roughly 12-16 weeks in? If yes, this is almost certainly your answer and the rest of the list is premature.
2. The vial is not what the label says
If you are at a genuine maintenance dose, have been there for weeks, and still feel nothing, the honest next question is whether the compound in the vial is the compound on the label at the mass on the label.
This is the point where third-party testing stops being a nice-to-have and becomes the only way to answer a question you cannot answer any other way. An underfilled vial — 60% of stated mass, say — produces exactly this presentation: a dose that should work, does not, and you have no way to see it. So does a vial containing a different compound entirely, or one with substantial impurity or degradation.
You cannot inspect for this. Clear solution, correct-looking lyophilised cake, sensible label — none of it carries information about content or mass. A certificate of analysis tied to your specific batch does, which is the entire reason COAs matter and why a vendor-supplied COA for a different lot is worth nothing. If you have no batch-matched testing, this branch of the differential is simply unresolved — and you can send a sample for independent testing rather than continue guessing.
How to tell if this is you: did the previous vial from a different source work at the same dose? A source-linked pattern is the strongest signal available without testing. If you have only ever used one source, you have no comparison and no way to distinguish this from the rest of the list.
3. The dosing math is wrong
This one hurts because it is entirely fixable and entirely invisible.
The failure mode is nearly always the same: confusing units with milligrams, or reconstituting at one concentration and dosing as though it were another. You drew 10 units. You believe 10 units is your dose. But 10 units of a vial reconstituted with 3ml of bacteriostatic water contains a fraction of what 10 units of the same vial reconstituted with 1ml contains — same vial, same units on the syringe, three times the difference in what enters you.
An order-of-magnitude error is the ugly version, and it is not rare. Somebody targeting 2.5mg who is actually delivering 0.25mg is at a tenth of their intended dose and cannot understand why the compound does nothing. From inside, that is indistinguishable from non-response. From outside, it is a decimal point.
How to tell if this is you: recompute from scratch. Not re-checking your previous arithmetic — recomputing independently, from vial mass and water volume, with the reconstitution calculator. Checking your own work reproduces your own error; the whole point is a path to the answer that does not run through the assumption you already made. If the calculator’s units do not match what you have been drawing, you have found it, and the fix is free.
4. The peptide degraded
Less common than the three above, and real.
Peptides in solution are not indefinitely stable. Reconstituted vials kept at room temperature for extended periods, repeated warm-cold cycling, a shipment that spent a week hot before arriving, or a vial pushed well past its useful window — all can reduce potency. The presentation is a compound that works progressively less well, or a specific vial that never worked while previous ones did.
Note the pattern this produces. Degradation usually presents as a vial-linked or timeline-linked failure, not a from-the-very-first-dose failure. If you have felt nothing from your first injection of your first fresh vial, storage is unlikely to be the story, and the storage guide will not have your answer. If effect dropped off after a heatwave or a shipping delay, it moves up the list considerably.
How to tell if this is you: is the failure specific to one vial or one period, with a plausible temperature event attached? Yes moves this up. Never-worked-from-day-one moves it down.
5. Genuine biological non-response
Now, finally, this.
It is real. Response to incretins varies widely between individuals, and the trial data shows it plainly — every published trial has participants at the bottom of the distribution who lost little despite full-dose, full-duration exposure. That tail is not a rounding artefact. Some people genuinely get much less from these compounds than the mean, and the reasons are not well characterised.
But look at what it takes to reach this conclusion honestly. You need to have been at a genuine maintenance dose, for months, with a batch-tested product, with independently verified dosing math, with sound storage — and still nothing. That is a demanding standard, and it is demanding on purpose, because every step you skip is a more likely explanation you left on the table.
And even the real thing is usually a matter of degree. Complete absence of any appetite effect at a full maintenance dose is rarer than a blunted effect. Most people described as non-responders in the trial data are not at zero; they are low in the distribution. Total nothing, in someone who has cleared items one through four, is unusual enough to be worth another look at items one through four.
Frequently Asked Questions
How long before I can say a GLP-1 is not working for me?
Not before you have reached the studied maintenance range and held it for several weeks — which for most titration schedules is roughly 12-16 weeks from the first injection, and often longer. The trials ran 72 weeks and the weight curves were still descending well past the midpoint. Concluding non-response at week four on a starting dose is calling a study from its first data point.
Why do I feel nothing on the starting dose?
Because the starting dose is not a therapeutic dose. Semaglutide starts at 0.25mg against a studied maintenance of 1.7-2.4mg; tirzepatide starts at 2.5mg against 5-15mg. The low steps exist so your gut adapts to escalation, not to produce an effect. Feeling nothing there is the expected observation, not a warning sign.
Could my vial be underdosed or the wrong compound?
It is possible and you cannot determine it by looking — appearance carries no information about content or mass. The only way to answer it is a certificate of analysis matched to your specific batch, or sending a sample for independent third-party testing. Without one of those, this branch stays unresolved, which is exactly why it is worth resolving before concluding anything about your biology.
How do I check my dosing math without repeating my own mistake?
Recompute from scratch rather than re-checking your arithmetic, because re-checking reproduces the original error. Start from vial mass and the volume of bacteriostatic water you added, run it through a calculator independently, and compare the resulting syringe units against what you have actually been drawing. A mismatch — especially a tenfold one — is the whole explanation.
Is being a true GLP-1 non-responder common?
Response varies widely and the trial distributions have a genuine low tail, so it is real. It is also rare relative to the alternatives, and complete absence of any effect at a full maintenance dose is rarer still than a blunted response. It is a conclusion to reach after underdosing, product quality, dosing math and storage have each been ruled out — not before.
Related reading
- Stalled on a GLP-1? Why You Plateau & How to Restart — the other problem, when it worked and then stopped
- Understanding Peptide Purity & COAs: How to Verify Quality — the only way to close item two
- Peptide Dosing 101: Concentration, Volume & Insulin Units — where the tenfold errors come from
- Peptide Storage Guide: How to Store Research Peptides — when a specific vial stops behaving