← All articles

GLP-1 Weight-Loss Plateaus: What the Trials Show

Bookmark this page

Save this article in your browser so you can find it again:

  • Computer: Press Ctrl+D (Windows) or ⌘+D (Mac).
  • iPhone or iPad: In Safari, tap Share, then Add Bookmark.
  • Android: In Chrome, open the menu and tap the star.

Other browsers may use a different menu.

Weight loss can slow and level off during continued GLP-1-based treatment. A plateau does not, by itself, show that a medicine has stopped having an effect. The useful questions are how researchers defined the plateau, who entered the analysis and what happened during follow-up. Semaglutide and tirzepatide studies illustrate why a group curve cannot supply an individual deadline. STEP 5 report; tirzepatide plateau analysis.

Publisher disclosure: Bio Research LLC operates Bio Peptides Labs and sells research materials. This commercially interested, AI-assisted article is not an independent scientific review. Selected original reports and indexed abstracts were checked November 14, 2026; access limits appear below. Clinical-trial findings do not establish equivalence of research catalog materials to prescription products.

Interpreting a weight-loss plateau requires its study definition, observation window and participant selection criteria.
An editorial reading framework for weight trajectories. It does not predict an individual's plateau or recommend a dose change.

First ask what the word “plateau” means

An unchanged measurement over a short interval, a slowing average and a research-defined plateau are different observations. A paper needs a numerical rule and a time window before its “time to plateau” can be interpreted. Otherwise, two studies might appear to disagree simply because they measured different things.

Research evidence map

Reading question

Example from the evidence

What to avoid concluding

Is this a group curve or an individual classification?

STEP 5 describes the average semaglutide weight trajectory

Every participant leveled off on the same date

Is a threshold specified?

The tirzepatide analysis used sustained 12-week intervals

Plateau necessarily means zero further change

Who qualified for analysis?

Some analyses selected adherent participants or responders

Results describe everyone who starts treatment

Is early response the same as final response?

A separate SURMOUNT-1 analysis followed early and late responders

One early measurement establishes the final outcome

Sources: STEP 5, plateau methods and early-response analysis.

What the tirzepatide plateau analysis found

The 2025 analysis of SURMOUNT-1 and SURMOUNT-4 was exploratory and post hoc. It included participants who adhered to tirzepatide and achieved at least 5% weight loss by the primary endpoint. Plateau meant less than 5% weight change over a 12-week interval and all subsequent 12-week intervals. The analysis included 1,438 selected participants from SURMOUNT-1 and 259 from SURMOUNT-4. In SURMOUNT-1, median times ranged from 24.3 to 36.1 weeks across baseline BMI categories. Original abstract.

Those are conditional results from selected participants. They cannot establish a universal deadline, nor does the definition require an absolutely flat scale. Because the classification depends on later intervals too, it is a retrospective description rather than a diagnosis someone can make from today's measurement.

What semaglutide's longer follow-up adds

STEP 5 randomized 304 adults with overweight or obesity, without diabetes. Its mean semaglutide weight curve leveled off around week 60 and remained broadly stable through week 104. Mean weight change at week 104 was −15.2% with semaglutide and −2.6% with placebo, both with behavioral support. Those figures concern change from the study's original baseline. Gastrointestinal adverse events were reported in 82.2% of the semaglutide group and 53.9% of the placebo group; the abstract describes most as mild to moderate. Those percentages are for this trial, not a forecast for another population. Original STEP 5 publication.

A stable lower average is different from a return to the starting weight. This is one reason a slowing curve should not automatically be called treatment failure. It also does not prove that every health outcome remains improved: each endpoint needs its own evidence.

A slow beginning is a separate question

A different post-hoc SURMOUNT-1 analysis included participants with substantial treatment adherence and weight measurements at four specified visits. Among its 278 participants below 5% weight loss at week 12, 250 reached at least 5% by week 72. Original early-response study.

This finding distinguishes an early response category from a later outcome. It does not promise that any particular person will eventually respond, and it does not establish how long someone should continue a medicine. People excluded by the analysis criteria may have different experiences.

Read the curve together with its analysis rules

A trial graph should show how many people contributed measurements at each visit. If participants stop treatment, miss visits or begin another intervention, the statistical analysis must decide how those events enter its estimate. The answer depends partly on the question the trial intends to address.

For example, STEP 5 distinguished an analysis including outcomes regardless of treatment discontinuation from one addressing continued treatment without rescue intervention. These are different research questions, not interchangeable versions of a headline percentage. STEP 5 analysis definitions.

For a reading worksheet, record the starting time point, analysis population, missing-data method and interval around the estimate. A mean with uncertainty is more informative than an isolated “expected loss” number. Our guide to reading a peptide study explains those distinctions.

Questions future plateau research should answer

We would like studies to distinguish sustained changes in weight from changes in body composition, physical function and daily life. A flat weight curve could coexist with changes in other measurements; that possibility needs measurement rather than assumption. See our separate muscle-preservation research agenda.

Useful future reports would publish their plateau definition in advance, show individual trajectories alongside group averages and account for withdrawals. They would also test proposed responses to a plateau against suitable controls. These are editorial research priorities, not trials conducted by Bio Peptides Labs or recommendations to add another compound.

Frequently asked questions

Does a plateau mean the medicine no longer works?

Not necessarily. Maintaining a lower weight and producing further weight loss are different outcomes. A weight curve alone cannot determine all effects of treatment.

Is there one normal month when everyone plateaus?

No universal timetable is established by these studies. Their populations, definitions, follow-up and analysis rules differ.

Does a plateau justify changing treatment or adding a peptide?

These analyses do not establish an individual treatment change or combination strategy. Personal decisions require a qualified clinician who can assess the broader clinical picture. This article provides no adjustment protocol.

Review method and source access

Checked November 14, 2026. We reviewed the plateau and early-response abstracts, the STEP 5 abstract through NCBI, and indexed STEP 5 full-text excerpts covering results, discussion and analysis definitions. Direct full-text access was intermittent; this was a selected-source review, not a systematic review. Weight regain after withdrawal is addressed separately in our stopping-GLP-1 article.

References

On November 20, 2026, we rechecked the three indexed abstracts and their reported populations and endpoints. Direct access to the STEP 5 full text was unavailable in this check; its previously reviewed indexed excerpts retain their November 14 access scope. These selected checks are not a systematic review.