The bottleneck is a category. The variable is what changes.
Knowing which area is limiting your results is the first half of a diagnosis. The variable inside that area is the part that actually needs to move, and it is not the same thing.
For educational and research purposes only. This is not medical advice and is not for human use guidance. Consult a qualified professional before acting on anything here.
Start here
Why the bottleneck is only half the diagnosis
Running the Protocol Bottleneck Tool tells you which of seven areas is most likely limiting your results. That is useful and it is not the full answer. The bottleneck is a category. The variable inside it is the thing that actually needs to change.
Most researchers stop at the category level. They identify the bottleneck, then adjust dose or timing without isolating which specific variable inside that bottleneck is the controlling one. Sometimes that works. Often it does not, because the adjustment was aimed at the right area and the wrong thing inside it.
A category tells you where to look. A variable tells you what to change. Adjusting a category is guessing inside a smaller box.
The variable moves even when the bottleneck does not
The variable inside a bottleneck also shifts over time. A protocol that was stalling on absorption six weeks ago may now be stalling because receptors have adapted to the compound, which means the bottleneck reads the same while the thing controlling it has changed underneath. If the diagnosis does not keep pace with the protocol, the adjustments fall behind it. Research suggests this is one of the more common reasons a protocol loses momentum after a strong start.
That is also the argument for working through this more than once. A read taken today describes today. The value comes from checking whether it still holds in four to six weeks, which is what part four of this guide is built for.
The seven bottlenecks and what sits inside them
Each of the seven covers a different part of the chain between a compound entering the body and a result showing up. Select one to see the variables it contains and how quickly they tend to move.
The speed reading matters for sequencing. A variable that can move within days is worth re-checking before one that moves over months, because a read taken against a fast variable goes stale sooner. It is not a ranking of importance.
A single clear variable, not a list of possibilities. If you finish a section with three equally plausible candidates, that is a signal in itself. It usually means data is missing rather than that the framework failed, and part two is where that gets separated out.
If you have not run the Protocol Bottleneck Tool yet, start there. It is free, takes about three minutes, and gives you the category this guide then works inside. You can also work through the parts below without it, since each one names which bottleneck it belongs to.
Free — no payment required
Read the rest
One email opens this guide and every other free research guide on the site. If you have already given your email anywhere on Project Theo, this will not appear again.
For educational and research purposes only. Not medical advice. No human use guidance. Used only for research guide access and occasional protocol updates from Project Theo.
Part two
What you can diagnose with the data you have
Four of the seven bottlenecks cannot be diagnosed from symptoms alone. They open by asking whether the data exists, and when it does not, the missing data is itself the finding. That is not a delay before the real answer. It is the answer, and it is the cheapest one to act on.
Check what you actually have rather than what you intend to get. The read below changes as you do.
Why missing data is a finding rather than a gap
There is a real difference between having measured something and found nothing unusual, and never having measured it. The first narrows the field. The second leaves it open while feeling like progress. When four of seven bottlenecks are gated on data you do not have, the honest read is that the field has not actually narrowed yet.
Research suggests the more expensive mistake is adjusting a compound to compensate for something that was never measured, because it adds a variable to a picture that was already unclear. Closing a data gap costs a test. Adjusting blind costs the ability to interpret everything that follows.
Part three
Four questions that isolate the variable
These questions do not return a protocol. They return the variable worth looking at next, which is a different thing and the reason this is a framework rather than a recommendation. Answer them from what you have observed rather than what you expected to see. Your answers save on this device.
The order is deliberate. Stress load and sleep come first because they are the cheapest to rule out and they do not require adding anything. A compound question asked on top of an unresolved stress load cannot be read cleanly either way.
Has anything significant changed in your daily life in the weeks leading up to the stall? Work pressure, poor sleep, travel, illness, or a sustained caloric deficit running longer than eight to ten weeks all raise cortisol, which is the hormone the body releases under sustained load. Research suggests even moderate sustained elevation is enough to blunt growth hormone output, blood sugar regulation, and anabolic signaling at the same time.
Did you have a clear initial response that has since faded, without any protocol change of your own? A strong early response that gradually diminishes while everything else stayed the same is the most consistent signal of receptor adaptation, meaning the proteins the compound binds to have become less responsive to the same signal.
Have you noticed injection site reactions beyond mild redness that fades quickly, or any uncertainty in how the compound has been mixed and stored? Persistent lumps, hardening under the skin, or localized swelling suggest the compound is pooling rather than absorbing. Incorrect water ratios when reconstituting, heat, light, or using a compound long after mixing all degrade it before it reaches circulation.
When did the stall appear relative to the last time you added or changed a compound, or shifted when you take one? If the timing overlaps, the change is the most likely variable. This one gets missed regularly, because a new compound is assumed to be helping and the possibility that it is competing for the same pathway does not get considered.
Part four
Checking whether the variable moved
The argument this whole guide rests on is that the variable moves while the bottleneck reads the same. If that is true, a single read is a snapshot and its value is limited until there is a second one to compare it against.
Save a read now, come back in four to six weeks, and save another. Each entry is compared to the one before it. The entries stay on this device and nothing is transmitted.
What a read that does not change is telling you
Two identical entries six weeks apart are information rather than a null result. Either the variable was correctly identified and has not been addressed yet, or it sits somewhere these four questions do not reach. Both are worth knowing and they point in different directions.
Nothing here resolves an individual protocol. It narrows which variable is worth looking at next, without knowing your bloodwork, your training load, or your history. That limit is real and worth stating plainly rather than working around.
Part five
When self diagnosis reaches its limit
This guide takes you from a bottleneck category to a specific variable. For most researchers that narrows the field enough to identify a productive next step. There are three situations where it does not, and they are worth naming rather than treating as a failure to work through it properly.
Three places the framework runs out
The first is when multiple variables appear active at once. Sustained stress raises cortisol, cortisol degrades blood sugar regulation, and poor blood sugar regulation caps metabolic output, so one input shows up as findings in three separate bottlenecks. Working through them individually returns three answers that are all correct and do not resolve into a single starting point.
The second is missing data, which part two covers. The third is when every section produces a plausible answer and none clearly leads, which usually means the reads are being generated by something upstream of all of them.
Reviewing one bottleneck at a time cannot see how two of them are affecting each other. That is a limit of the method, not of the effort.
Why the interaction is the hard part
Each part of this guide asks about one area in isolation, which is what makes it workable on your own. It is also what makes cross bottleneck interaction invisible to it. A variable in Bottleneck 07 that is producing findings in 01 and 04 will read as three independent problems, and adjusting for each separately treats symptoms of a single cause as three causes.
When several variables are genuinely active, the useful output is the order to address them in. That priority sequence is what a full protocol review surfaces and a section by section read cannot. The most reliable place to start, though, is upstream of all of this, with the free tool that names your category in the first place.
Take the PDF with you
The original diagnostic guide as a PDF. The page is the current version and reflects the framework as it stands now, including several claims that have since been reframed.
Download the PDFNext step
Find your bottleneck first
The Protocol Bottleneck Tool maps your exact situation across the seven bottleneck categories and tells you where the friction most likely is. It does not ask which compound you are on. It reads the pattern first, then points you to the category this guide works inside. Free, no purchase, and about three minutes.