Fat loss was working, and then the scale stopped
When a GLP-1 protocol slows, the reflex is to raise the dose or add a compound. Often that treats a problem that is no longer the problem. The same stall has a different cause depending on where the protocol is. This guide is a way to work out which cause you are actually looking at, and what the research does and does not support.
For educational and research purposes only. This is not medical advice and is not human-use guidance. Consult a qualified professional before acting on anything here.
Part one
Why the same stall has different causes
Most people who reach a stall do the same three things. Adjust the dose, add a compound, switch to something stronger. Results do not improve and sometimes they get worse. The reason is usually not the compound. It is that the question being answered is the wrong question for where the protocol is.
Fat loss does not fail because a compound stops working. It moves through phases, and at each phase the variable that is actually holding results back shifts. A GLP-1 compound suppresses appetite, which is exactly what is needed when intake is the limiting variable. But appetite suppression does not raise energy output, restore efficiency, or improve recovery. Push it harder once the protocol has moved on, and you are adding pressure to a variable that is already handled.
That is why identifying the phase comes before any change. The explorer below walks the four phases and what limits each one. It is a lens for reading your own situation, not a verdict on it.
The body has not yet adapted to eating less, so how much is going in is what controls results. A GLP-1 compound is the rational tool here because it acts directly on intake. Adding metabolic or growth hormone compounds at this stage adds complexity without addressing anything the GLP-1 is not already handling. The task that matters most in Phase 1 is protecting lean mass from about week four onward.
Worth pausing on. Most researchers feel confident they have identified their phase, and in the broad sense they are usually right. The detail is where it slips. Phase 2 and Phase 3 overlap more than a clean framework suggests, because cortisol, training volume, and sustained restriction can push a protocol into Phase 3 territory while the scale still looks like Phase 2. If sleep has declined, training performance has dropped, or fatigue is present on rest days alongside the stall, that overlap is worth ruling out before reading anything as a pure Phase 2 problem.
Free — no payment required
The rest of the guide
Part one is the lens. The foundation check, the plateau sequence, the compound reference, and where the framework stops are below. Enter your email to read the rest. One step, no account needed.
No spam. One email unlocks this and the other free research guides on the site.
Part two
The foundation check, before any compound
Before any compound is considered for any stall, four inputs need checking. Research consistently shows that compounds added on top of an unstable foundation produce a fraction of their documented effect. This is not a soft suggestion, it is a filter. If one of these is genuinely compromised, resolving it is the first move, not a compound.
One principle runs underneath all of it. Peptides amplify. They do not replace infrastructure. If sleep is poor, stimulants are high, stress is unmanaged, and nutrition is chaotic, every compound amplifies the chaos rather than correcting it. The foundation comes first because everything downstream is multiplied by it.
Nothing checked yet. Work through the four inputs above and a read will appear here.
Part three
Reading a Phase 2 plateau
The body adapts. After weeks of eating less, metabolic rate settles lower, and what was a real deficit may no longer be one. That is a normal defense mechanism, not a compound failing. GLP-1 compounds do not stop working at a Phase 2 stall. They stop being enough on their own, because the intake role is already fulfilled and the variable has moved to output. Pushing intake suppression further at this point tends to cut energy faster than it moves fat.
So the scale stopping is not evidence the dose is too low. For most people at this stage it is evidence the mechanism has done its job and the next variable needs attention. The sequence below runs three checks in the order that rules out the cheapest explanation first. Answer them in order.
Intake precision, before anything else. Metabolic rate adapts across an extended GLP-1 protocol, so a deficit that was around five hundred calories at week four can be closer to a hundred by week sixteen without anything visibly changing. Research suggests most stalls at this stage resolve with a small intake correction, not a dose increase. This is the cheapest variable to rule out, and it settles the read on its own.
Ask the search engine about intake driftThe recent escalation itself, not a missing compound. A dose step that cuts energy without producing proportional fat loss points to the dose being past its useful point rather than short of it. The variable here is the escalation, and it is worth reading alongside the seven day intake check rather than treated as a reason to add something.
Ask the search engine about escalation and energyIf intake is controlled, energy holds, and the compound is at full titration, the intake mechanism looks fully used. That makes the next question an output and mechanism question rather than an intake one, and the nature of it depends on which compound is running, since a single receptor stall and a three receptor stall are different problems. That is a diagnostic read rather than something this framework resolves from the outside. Part four sets out what each compound does and what the evidence shows, and the diagnostic tool works through the specific compound.
Open the Compound DiagnosticIf any of the three is uncertain, that uncertainty is the variable. A stall cannot be attributed cleanly while intake, energy, or titration is an open question, so the cheaper move is to make one of them certain first. The search engine covers each of these individually.
Ask the search enginePart four
The compounds, by what they do and what the evidence shows
This is a reference, not a running order. Each compound below answers two questions kept apart. What it does, in plain language. And separately, what human trials have measured, because a clean mechanism and strong human evidence do not always travel together. Nothing here is a dose, a sequence, or a recommendation to add or withhold anything.
The GLP-1 receptor class
These three are separated mainly by how many receptors they engage, not by the dose of any single one. The first two work almost entirely on intake, meaning how much you eat. The third adds a mechanism the other two do not have.
| Compound | What it does | What the evidence shows |
|---|---|---|
| Semaglutide Approved |
Acts on one receptor, GLP-1, a gut hormone that signals fullness and slows how fast the stomach empties. A pure intake mechanism. It changes how much you want to eat, not how much energy you burn. | Around fifteen percent average body weight loss over roughly sixty-eight weeks in the STEP trials. Approved for weight management, so the human evidence is about as settled as it gets in this area. |
| Tirzepatide Approved |
Adds a second receptor, GIP, on top of GLP-1. Still primarily an intake mechanism, but engaging two receptors produces stronger results than one when the two are compared head to head. | Roughly twenty to twenty-two percent at the top dose in the SURMOUNT trials, ahead of semaglutide in direct comparison. Also approved, also backed by large trials. |
| Retatrutide Investigational |
Adds a third receptor, glucagon, alongside GIP and GLP-1. The glucagon receptor is the real difference. It works on output, signalling the body to burn stored fuel and oxidize liver fat rather than only lowering intake. That is why it behaves differently, not simply more strongly. | About twenty-eight percent at eighty weeks in the pivotal TRIUMPH-1 trial reported in 2026, the largest figure yet in this class. Still investigational and not approved anywhere as of mid-2026, with a filing expected around late 2026. Trials also flagged an abnormal skin sensation, tingling or prickling, in up to about a fifth of people at the highest dose. |
The amylin pathway
A different appetite signal entirely, rather than pushing the GLP-1 system harder.
| Compound | What it does | What the evidence shows |
|---|---|---|
| Cagrilintide Investigational |
Works through amylin, a hormone released with insulin after meals that signals fullness through a different receptor system than GLP-1. So it stacks a second, independent fullness signal rather than turning up the same one. | In the phase 3 REDEFINE trials, the semaglutide combination reached about twenty-three percent against about sixteen for semaglutide alone. On its own, cagrilintide was actually weaker than semaglutide alone, around twelve percent. The value showed up in the combination, which also missed its own target and did not beat tirzepatide in a separate trial. Not yet approved. |
The fat mobilization compounds
This is where the gap between mechanism and evidence is widest. Each has a clean, appealing mechanism story. The human evidence ranges from thin, to absent, to actively negative. Read the mechanism and the evidence as separate columns.
| Compound | What it does | What the evidence shows |
|---|---|---|
| HGH Fragment 176-191 Preclinical only |
The tail end of the growth hormone molecule, the piece tied to fat breakdown, isolated to trigger lipolysis, meaning fat release from cells, without the blood sugar and growth effects of full growth hormone. | The fat release effect is real in cell and rodent studies, but the fragment itself has essentially never been properly tested in humans. Nearly all human relevant data comes from its modified cousin AOD-9604, not from the fragment. Its reputation as a dependable fat loss tool runs well ahead of what has been shown in people. |
| AOD-9604 Failed in trials |
A stabilized version of the 176-191 fragment, same fat release mechanism, developed specifically as an anti-obesity drug candidate. | Often presented as having clinical data supporting fat loss. The larger evidence points the other way. Early small studies hinted at modest results, but the more definitive phase 2b trial failed to beat placebo on its main measure, the obesity program was discontinued, and it was redirected toward joint research. Never approved for fat loss. |
| AICAR Preclinical only |
Switches on AMPK, the cell's energy sensing switch that flips on when fuel runs low and pushes cells to burn energy. A 2008 mouse study nicknamed it exercise in a pill. | The AMPK mechanism is real, but the human fat loss evidence is essentially absent, and the striking results are from rodents. It also sits on the anti-doping banned list as a metabolic modulator, which is most of why the name circulates. |
| 5-Amino-1MQ Preclinical only |
Blocks NNMT, an enzyme inside fat cells that acts like a metabolic brake and becomes more active during prolonged restriction. Blocking it is proposed to release that brake. | The idea is genuinely interesting, but the data is preclinical, mouse and cell studies. There are no human trials establishing that it produces fat loss in people, and descriptions of a gradual effect over a set number of weeks describe a usage pattern trials have not tested. |
Mechanism and evidence are two separate axes. The most appealing fat loss mechanism is often the one with the least human evidence, and in at least one case the larger trial was negative. Holding them apart is most of what separates a careful read from a hopeful one.
Part five
Where the framework stops
The path in this guide works for most stalls that are what they look like. A single phase, a single bottleneck, one clear limiting variable. Run the checks in order and most plateaus at this stage either resolve or at least become legible.
There is a category it cannot reach. Not a harder version of Phase 2, but a situation where two bottlenecks are active at the same time, or where the phase looked right while a second variable ran underneath it. Address one problem in that situation and the other often gets worse, and the conclusion becomes that the compounds do not work when the actual issue was a misread. That pattern needs more resolution than any general framework can give from the outside, because it depends on the specific interaction between compound, phase, recovery, and how the protocol has adapted.
A framework reads the common case. A two variable interaction needs a read built from the actual inputs.
That is the line where a general guide stops and a situation-specific read picks up. Mapping which of the bottleneck categories is actually the limiting one, against your real inputs rather than the common case, is what narrows a two variable stall back down to a single variable to examine.
Next step
When the read is more than one variable
If you have run the checks and the stall still is not legible, that is usually the two variable case a general framework cannot reach from the outside. The Protocol Bottleneck Tool maps your situation across the bottleneck categories and returns where the friction most likely is. It is free and takes about three minutes.