The GLP-1 Weight-Loss Plateau: Why It Happens and What the Evidence Says to Do

A plateau while you're still taking a GLP-1 is a different problem from what happens after you stop one — this article is about the first, not the second. Weight loss slowing or stalling while treatment continues is common enough to ask about constantly, and the evidence on why is more specific than "your body adjusted."

The trial curves themselves show it

Even in the pivotal trials, weight loss on a GLP-1 doesn't proceed as a straight line. Semaglutide's STEP 1 trial measured mean weight loss through 68 weeks, and tirzepatide's SURMOUNT-1 trial measured its own primary endpoint through 72 weeks¹² — both trials were specifically designed to run long enough to capture the point where the average participant's weight-loss curve begins to flatten, rather than continuing to fall indefinitely. That flattening in trial data is the same phenomenon patients describe as a plateau — it isn't a deviation from what the drug does, it's consistent with how the drug's effect behaves over time at a given dose, in the aggregate data these trials were built to capture.

Not everyone plateaus at the same point — and that's now measurable

One of the more useful recent findings is that response variability isn't random noise — it has an identifiable genetic component. A large genetic-predictor study published in Nature analyzed genetic variants associated with the degree of weight loss and side effects experienced on GLP-1 receptor agonists³, providing real evidence that some of the person-to-person difference in how much weight loss a given dose produces — and, by extension, where a plateau tends to occur — is written into individual biology rather than being a matter of effort or adherence. That's a meaningfully different and more accurate frame than treating a plateau as something the patient did wrong.

What clinicians actually have to work with at a plateau

A review of obesity pharmacotherapy specifically addressing how to select, tailor, and combine treatment describes the real toolkit available once a plateau is reached: dose optimization within the labeled titration ceiling, switching to a different mechanism, or combination approaches, matched to the individual patient rather than applied as a blanket next step⁴. In practice, that usually means one of three conversations with a prescriber: confirming you've actually reached your labeled maintenance dose rather than plateauing on a sub-therapeutic titration step (see the full titration schedule to check), discussing whether a molecule switch makes sense (see switching from semaglutide to tirzepatide for what that evidence shows), or addressing non-pharmacologic factors — protein intake, resistance training, sleep — that influence the plateau independent of the drug itself.

What a plateau is not

A plateau while still taking the medication at an adequate dose is a different situation from regain after stopping treatment entirely — the physiology and the fix are not the same, and conflating the two leads people toward the wrong next step. See what actually happens when you stop a GLP-1 if what you're experiencing is regain after discontinuing, rather than a stall while still on treatment. A plateau is also not proof the drug has "stopped working" in a way that requires immediately abandoning it — it's a well-documented feature of the trial data itself, and the honest next step is usually a conversation about dose, molecule, or lifestyle factors with your prescriber, not an assumption that you need to switch providers or give up.

The bottom line

If your weight loss has slowed while you're still on treatment at your labeled maintenance dose, that's consistent with how these drugs behave in their own pivotal trial data, and increasingly explainable by real, individual biological variation rather than something you're doing wrong. Bring the specific timeline and your current dose to your prescriber, and price out whether a molecule switch is worth exploring using the cost calculators before committing to a change. The trial and genetic-predictor research behind this is indexed in our research library. Not medical advice — see our medical disclaimer and talk to your prescriber before changing your dose or molecule.