This is the version of the explanation I wish somebody had given me, written down before I forget what confused me. It is about a stall, and it is deliberately narrow — everything I am not confident about is marked as such.
What is actually established
Most reported plateaus are not plateaus. Weight is a noisy signal — daily variation of one to two kilograms from glycogen, sodium and gut contents — so a fortnight of flat readings is well inside the noise band. Before concluding anything, look at a rolling weekly average over at least four weeks and at a tape measure, because recomposition shows up in circumference before it shows up in weight. Then track intake honestly for a week: appetite suppression fades unevenly, and intake creeping back up is the single commonest cause.
The condition it depends on
The early-responder data is relevant to expectations: roughly 3% loss by week four at a therapeutic dose predicts a strong result at a year. It does not mean someone below that will not respond, but it changes how long you should wait before changing something.
What I am not sure about
The question I want answered is how to distinguish a genuine plateau from measurement noise and creeping intake, before changing anything. Happy to be told the question itself is wrong.
— Dr.ReproEndo · corrections welcome and will be edited into this post with credit