A clinical trial can describe average weight change in a study group. It cannot tell you exactly how much weight you will lose. The useful starting point is a named study, its participants, comparison group and measurement week.
What selected trials actually found
These studies tested specified injectable regimens under trial conditions. Read each row as a separate study. The first two trials compared a medicine with placebo; the third directly compared the two medicines.
| Study | Participants and design | Time point | Reported mean reduction | Context and source |
|---|---|---|---|---|
| STEP 1 | 1,961 adults with obesity, or overweight plus a weight-related condition; without diabetes. Randomized, double-blind. | 68 weeks | Semaglutide study arm: 14.9%. Placebo: 2.4%. |
Studied semaglutide 2.4 mg weekly. Both groups received lifestyle intervention. Wilding et al., 2021. |
| SURMOUNT-1 | 2,539 adults with obesity, or overweight plus a weight-related complication; without diabetes. Randomized, double-blind. | 72 weeks | Tirzepatide study arms: 5 mg: 15.0%. 10 mg: 19.5%. 15 mg: 20.9%. Placebo: 3.1%. |
Three assigned weekly regimens; values use the treatment-regimen estimand. Jastreboff et al., 2022. |
| SURMOUNT-5 | 751 adults; eligible participants had obesity, or overweight with a weight-related condition, without type 2 diabetes. Randomized, open-label. | 72 weeks | Tirzepatide: 20.2%. Semaglutide: 13.7%. |
Maximum tolerated weekly doses: tirzepatide 10 or 15 mg; semaglutide 1.7 or 2.4 mg. Least-squares means. Published online May 11, 2025. Aronne et al., 2025; study protocol (PDF). |
Reductions are shown as positive percentages for readability; the papers report negative percentage changes. This table describes study regimens, not dosing instructions. It is a selected overview, not a review of every trial or formulation.
What an average can and cannot tell you
A mean combines outcomes across a group. It does not describe every participant, and it is different from the percentage of participants who reached a threshold.
For example, STEP 1’s week-68 responder results reported 86.4% for at least 5% weight reduction and 50.5% for at least 15% in the semaglutide group. Those are shares of participants meeting different thresholds, not two estimates of average weight change. See the original report for the analysis and denominators.
Uncertainty matters too. In SURMOUNT-5, the 95% confidence intervals for mean percentage change were −21.4 to −19.1 with tirzepatide and −14.9 to −12.6 with semaglutide. These intervals concern group estimates, not the range of individual experiences. Read the study abstract.
Keep the time point and study support attached
The 68- and 72-week results above cover more than a year. Dividing a final average by the number of weeks does not turn it into a reliable weekly target.
STEP 1 included diet and activity counselling alongside study treatment. That support is part of the study context. Its results do not establish which services a particular telehealth plan provides. Check a plan’s actual written follow-up arrangements instead of assuming they match a trial.
What a withdrawal study tells us
The STEP 1 extension followed a subset of participants after both study medication and lifestyle intervention ended at week 68. Its exploratory analyses included 327 people.
In the semaglutide group of that subset, mean weight reduction had been 17.3% at week 68. By week 120, participants had regained 11.6 percentage points on average, leaving a mean reduction of 5.6% from the original starting weight.
Those figures describe this selected follow-up group. They are not the original trial’s full-population average and do not establish an inevitable outcome for every person. Because lifestyle intervention also ended, the study should not be described as isolating medication withdrawal alone.
Check whether a claim matches its evidence
Before treating a website’s outcome percentage as useful evidence, look for the study or dataset, exact product, population, follow-up time, missing-data explanation and harms. If those details are absent, record the gap rather than filling it with assumptions.
Trials of a specified product do not establish equivalent results for every preparation sharing an ingredient name. FDA explains that compounded drugs are not FDA-approved and are not reviewed by the agency for safety, effectiveness or quality before marketing. Our compounded versus FDA-approved guide covers that distinction.
A trial summary cannot determine your eligibility or explain your own response. Questions about treatment, symptoms or changes in progress belong with a clinician who knows your history.
Sources and editorial scope
We checked the linked original research reports or their author-written abstracts and FDA information on September 14, 2026. STEP 1 was funded by Novo Nordisk; SURMOUNT-1 and SURMOUNT-5 were supported by Eli Lilly. Funding and study limitations should be read with the findings.
We have not conducted these trials, verified individual patient outcomes or clinically reviewed this article. Historical study results are identified by their publication years. This selected overview cannot replace an assessment of the full evidence for a clinical decision.