# What a Plateau Actually Means

By Nora Bissett, Pricing Editor (https://glpwatchdog.org/authors/nora-bissett) | Published: 2026-08-20 | Updated: 2026-10-27

Canonical page: https://glpwatchdog.org/research/what-a-plateau-actually-means

> Nine in ten people had plateaued by week 72. Because of how the researchers defined a plateau, you can be counted as plateaued while still losing about 16% of your body weight a year.

## Key points

- Median time to plateau: 24.3 weeks (overweight) to 36.1 weeks (class II–III obesity).
- 'Plateau' = under 5% change per 12 weeks. That is compatible with losing ~16% a year.
- Higher doses, younger age and female sex all pushed the plateau later.
- Only adherent participants who lost ≥5% were included, so this is the shape of SUCCESSFUL treatment.
- Post-hoc, and the paper carries a published correction.

Nine in ten people in this analysis had reached a weight plateau by week 72.[1] That sounds like weight loss stopping for almost everyone, but that is not what the number means. Because of how the researchers defined a plateau, **you can be counted as plateaued while still losing about 16% of your body weight a year**.

## The definition, which is in the methods

A participant was classified as having plateaued at the start of the first 12-week interval in which their weight changed by **less than 5%**, provided every subsequent 12-week interval also stayed under that threshold.

> Losing 4% every twelve weeks counts as a plateau.

Four percent per quarter compounds to roughly 16% over a year. For someone starting at 100 kg that is around 16 kg in the year after they were declared to have plateaued. This is a threshold for *the rate slowing*, not a description of weight loss ending.

> **A definition like this is how a fuzzy idea becomes measurable.** The researchers needed a rule that could be applied to every participant identically, and this one is reasonable. The problem arrives when the resulting statistic travels without it, and “90% plateaued by week 72” is exactly the kind of sentence that travels alone.

A separate systematic review shows how much the definition matters. It asked the same question for a surgical purpose, defined a plateau as under 3% loss over at least a month, and arrived at **53 weeks** rather than 24 to 36. A stricter threshold for the same phenomenon made half a year of difference (see [when is it safe to have the skin removed](https://glpwatchdog.org/research/when-is-it-safe-to-have-the-skin-removed)).

## When it happened

*Median time to weight plateau in SURMOUNT-1, by baseline BMI category. 1,438 participants.[1]*

| Starting category | Median time to plateau | Plateaued by week 72 |
| --- | --- | --- |
| Overweight | 24.3 weeks | 90.2% |
| Obesity class I | 26.0 weeks | 88.9% |
| Obesity class II | 36.1 weeks | 87.6% |
| Obesity class III | 36.1 weeks | 87.8% |

The pattern is orderly: **the heavier someone started, the longer their rapid phase lasted**. It ran about six months for people in the overweight category against roughly eight and a half for the heaviest, a statistically significant difference. Results in the second trial, SURMOUNT-4, were similar.

Three other factors pushed the plateau later: **higher tirzepatide doses (10 or 15 mg), younger age, and being female**. None of those is something to act on directly, but all of them are useful context for someone comparing their own curve to somebody else’s.

## Who is in this analysis

> **Only participants who stayed adherent *and* lost at least 5% of their weight were included.** Everyone who stopped the drug, and everyone who took it without responding, is absent. So this describes the shape of the curve *among people for whom the treatment worked*. That is useful if you are one of them, and says nothing about anyone else.

That selection also means the plateau timings cannot be read as a forecast before starting. They are a description of what successful courses of treatment looked like in retrospect. A plateau is also not a sign the drug has slowed your metabolism; [the measurements](https://glpwatchdog.org/research/does-ozempic-slow-your-metabolism) say otherwise.

## What to do with your own curve

The most common reason people conclude a drug has stopped working is that the scale moves less than it did in month two. Month two is the steepest part of a curve that flattens for everyone, so any later month compared against it will always look like failure. Adding a second drug at that point is common in practice and largely untested; [what we know about adding phentermine to a GLP-1](https://glpwatchdog.org/research/phentermine-and-a-glp1-together) sets out the evidence.

- **Expect the rapid phase to run roughly six to eight months**, longer if you started heavier or are on a higher dose.
- **Slowing is not stopping.** Under this analysis’s own definition, a plateau is compatible with continuing to lose a substantial amount.
- **The cardiovascular benefit does not track how fast you lose.** In SELECT, protection appeared across every weight category with no relationship to early weight loss (see [losing weight predicted more heart attacks](https://glpwatchdog.org/research/losing-weight-predicted-more-heart-attacks)).
- **Early response does carry information** about the year ahead, which is a separate question (see [losing fast early](https://glpwatchdog.org/research/losing-fast-early-what-it-predicts)).

One limitation on all of it: this is a post-hoc analysis, meaning the plateau question was asked of trial data collected to answer something else. The paper also carries a small published correction to a figure in its own abstract.

## Frequently asked questions

### When does weight loss plateau on tirzepatide?

Median time to plateau ranged from 24.3 weeks for people starting in the overweight category to 36.1 weeks for those with class II or III obesity. By week 72, roughly 88–90% had reached a plateau by the study's definition.

### Does a plateau mean weight loss has stopped?

Not under this definition. A plateau was recorded when weight changed by less than 5% over a 12-week interval, so someone losing 4% each quarter (around 16% a year) counts as plateaued.

### Why did heavier people take longer to plateau?

The analysis reports the association rather than explaining it. People starting with higher BMI had longer rapid phases: about six months in the overweight category against roughly eight and a half in the heaviest.

### What else made the plateau come later?

Higher tirzepatide doses of 10 or 15 mg, younger age, and female sex were all associated with a longer time to plateau.

### Does this describe everyone taking the drug?

No. Only participants who remained adherent and lost at least 5% of their weight were included, so it describes the shape of successful treatment rather than the average experience.

## References

1. Horn DB, Kahan S, Batterham RL, et al. Time to weight plateau with tirzepatide treatment in the SURMOUNT-1 and SURMOUNT-4 clinical trials. Clinical Obesity. 2025. PMID: 39800653. https://pubmed.ncbi.nlm.nih.gov/39800653/

---

Cite as: GLP Watchdog, Nora Bissett, "What a Plateau Actually Means", published 2026-08-20, updated 2026-10-27. https://glpwatchdog.org/research/what-a-plateau-actually-means

Canonical HTML page: https://glpwatchdog.org/research/what-a-plateau-actually-means

Medical disclaimer: Nothing here is medical advice. We report on this subject; we do not treat anyone. Whether to begin a medication, supplement, or treatment, whether to adjust it, and whether to stop are questions for a clinician licensed to answer them, who knows your history.
