Evidence review · from the 250-topic plan
The plateau audit: six causes to rule out before blaming the drug
“It stopped working” is a conclusion; a plateau is a measurement — and between them sits an audit most people skip. Run the six causes in order: (1) you're not at steady state yet (each rung needs its month); (2) intake drifted (appetite adapts; portions creep; liquids don't register); (3) the math converged (a lighter body burns less — every loss narrows the gap that causes loss, so slowing is arithmetic, not betrayal); (4) the scale is lying short-term (water, glycogen, salt, cycles — trends beat days); (5) you're recomposing (training adds muscle while fat leaves — the tape knows); (6) a true pharmacological ceiling — the only cause that's actually about the drug, and the one with named next moves. Audit in order, because causes one through five are cheaper than cause six's conclusions.
Cause 1: you're grading a dose that hasn't arrived
Multi-day half-lives mean each rung takes four-to-five weeks to reach full blood levels — so a “plateau” two weeks after a dose change is a preview being graded as a verdict. The fix is a calendar, not a decision: date your last rung change, add a month, audit then. (The mirror error also lives here: judging a hold as failure when it's the measurement window working.)
Cause 2: intake drifted while nobody watched
The drugs suppress appetite; they don't disable it, and adaptation plus confidence quietly reopens gaps: portions grow back toward old defaults, “just a taste” frequency rises, and liquid calories — lattes, juices, alcohol — slip past a fullness signal tuned to solids. The audit: ten honest days of logging, liquids included, before any conclusion. Nearly everyone who runs this step finds something; the finding isn't a character flaw, it's the cheapest fix on this page — and it's why the noise rating makes a useful co-metric: quiet noise with creeping intake is a habits problem, returning noise is a dose data point.
Cause 3: the math converged — the plateau that means it worked
Weight loss runs on an energy gap, and every kilogram lost shrinks it: a lighter body spends less at rest and in motion, so the same intake that once produced steep loss eventually produces balance. This is the least understood sentence in weight medicine: a plateau at a much lower weight is largely the arithmetic of success — the trials' own curves bend for exactly this reason, steep through month eight and flattening after, per STEP 1's shape. The response isn't outrage at the drug; it's a decision about whether the new equilibrium is home (welcome to maintenance) or a waypoint (see cause six's moves).
Cause 4: the scale is measuring water, salt, and Tuesday
Day-to-day weight swings by pounds on glycogen, sodium, hydration, hormones, and bowel timing — noise that dwarfs a week's true fat change. Measurement literacy: weigh under constant conditions, judge two-to-four-week trend lines only, and keep the tape and photos as slower, honester instruments. A “two-week stall” inside normal noise isn't a plateau; it's weather.
Cause 5: you're trading fat for muscle
Anyone running the resistance-and-protein program properly can see scale-stillness while composition improves — fat leaving, lean arriving, waistline shrinking through a frozen number. The audit: measurements, strength logs, and how clothes fit. This is the plateau to celebrate, and mislabeling it as failure has talked more than one patient out of the exact behavior protecting their result.
Cause 6: the real ceiling — and its named moves
If the audit clears — steady state reached, intake logged clean, trend flat over a month-plus, composition static — you've likely met your response ceiling at the current dose, and the moves have names: climb a rung if the ladder has room and tolerability allows (SURMOUNT-1's dose-response is the case for it); switch molecules if you're ceilinged on semaglutide's ladder — the head-to-head is the case, and the switch guide is the map; or declare maintenance and defend the result the withdrawal trials say needs defending. What the audit buys is standing: a dose or molecule conversation backed by a month of clean data is medicine; the same conversation backed by frustration is guessing. And one architecture note this site owes you: on dose-proof flat pricing, cause-six moves cost $0 extra — which keeps the decision clinical, exactly where it belongs. Ceiling moves that don't cost extra ↗
FAQ
Why did my weight loss stall on tirzepatide or semaglutide?
Audit six causes in order: dose not yet at steady state, quiet intake drift, the shrinking energy gap of a lighter body, short-term scale noise, muscle-for-fat recomposition — and only then a true dose-response ceiling with named next moves.
Is a plateau normal on GLP-1 medications?
Yes — the trials' own curves flatten after early months as the energy gap converges; a plateau at a much lower weight is largely the arithmetic of success.
What should I do after a confirmed plateau?
With a clean month of data: discuss climbing a rung, switching molecules if ceilinged on semaglutide, or declaring maintenance — clinician conversations that data upgrades from guessing to medicine.
Sources
- Trial weight-curve shapes (STEP, SURMOUNT) — the physiologic flattening.
- Energy-gap and metabolic-adaptation literature; body-composition substudies.
- Steady-state pharmacokinetics and dose-response data — companion analyses.