Hit a GLP-1 plateau? Here's what the trial curves actually show

Week 19: the scale hasn’t moved in three weeks. The forums offer two equally unhelpful diagnoses — “you’ve stalled, switch meds” and “trust the process” — with nothing to tell you which applies. There’s a better way: look at the published curve you’re implicitly comparing yourself to.

What the trial curves look like

Plot the average weight trajectory from STEP 1 (semaglutide 2.4 mg) or SURMOUNT-1 (tirzepatide) and you won’t see a straight line. You’ll see:

• A steep early phase through roughly the first half-year, where most of the total change happens;
• A long flattening — the average keeps drifting down, but slower every month;
• An eventual plateau, by design: endpoints landed around ~15% at 68 weeks (STEP 1) and up to ~21% at 72 weeks (SURMOUNT-1, highest dose).

Even within the steep phase, individual trajectories wobble — water shifts, hormonal cycles, and measurement noise produce flat stretches of two to four weeks that mean nothing. A three-week stall at week 19 is well inside normal trial behavior.

Plateau or noise? Three checks

1. Zoom out to a month. Compare your 4-week rolling average against itself a month earlier — not day against day. Daily weights are weather; averages are climate.
2. Check your week-vs-trial position. Are you broadly tracking the published trajectory’s shape at your week of treatment? The shape matters more than the absolute number — you may sit above or below the average the whole way and still be on a healthy curve.
3. Check the boring causes. Mid-titration (a level still building), recent missed or late shots, or a medication switch all flatten weeks legitimately.

If the rolling average has been truly flat for 6–8+ weeks at a stable full dose, that’s a real conversation to bring to your prescriber — with data, not vibes.

Benchmarks beat forums

The reason plateau panic thrives is that most people have no reference curve at all. ShotLock’s progress view plots your weight against the cited trial trajectory at your same week of treatment — mid-trial points estimated from published figures, endpoints exact, every number traceable in the Research Library. It won’t tell you you’re “ahead” or “behind” as a judgment; it shows you the context the forums can’t.

And on the weeks the scale refuses to cooperate, your progress photos usually have a different opinion worth hearing.

Trial figures describe study populations with lifestyle support, not individual predictions. Not medical advice.