STEP 2 found average weight loss of 9.6% with semaglutide 2.4 mg and 3.4% with placebo, an inactive comparison treatment. Those STEP 2 trial results cover 68 weeks in adults with type 2 diabetes. Average loss with 1.0 mg was 7.0%. Blood-sugar changes were much closer between the two semaglutide groups.

Who joined the STEP 2 semaglutide diabetes study?

Researchers asked whether semaglutide 2.4 mg produced more weight loss than placebo in people who also had type 2 diabetes. The lower semaglutide dose provided another comparison.

Melanie Davies and colleagues reported STEP 2 in The Lancet in 2021. It enrolled 1,210 adults across 12 countries. Participants had to be at least 18; their average age was 55.

BMI, a weight-to-height measure, had to be at least 27. Entry HbA1c, a blood test reflecting roughly three months of blood sugar, was 7–10%. People used diet and exercise alone or up to three stable diabetes medicines taken by mouth.

Current insulin use, severe kidney impairment and unstable diabetic eye disease were among the exclusions in the trial registry.

The three groups were assigned by chance: 404 people to 2.4 mg, 403 to 1.0 mg and 403 to placebo. Everyone received food and activity counseling.

The trial used a double-dummy design: everyone had two weekly shots to hide the different treatments. Semaglutide groups received one active and one inactive shot; the placebo group received two inactive shots. Doses increased gradually. Treatment lasted 68 weeks, followed by seven weeks without trial treatment.

What was measured, and how?

The two main measures were percentage weight change and the share losing at least 5% of starting weight. The main comparison was 2.4 mg against placebo.

The main analysis kept people in their assigned groups even if they stopped treatment or received extra weight-loss treatment. Missing measurements were estimated using follow-up data from people in the same group, including those who had stopped.

A second analysis estimated what would happen with continued assigned treatment and no extra weight-loss treatment. That is a different question, rather than a second result that cancels the first. Our estimand explainer covers how a study defines the treatment effect it wants to measure.

Diabetes medicines could also be adjusted when needed. The trial therefore did not test semaglutide while keeping every other medicine unchanged.

STEP 2 trial results: weight and blood sugar

The main analysis found 9.6% average weight loss with 2.4 mg versus 3.4% with placebo. The estimated difference was 6.2 percentage points. Against 1.0 mg, the estimated advantage was 2.7 points.

The paper's estimates at week 68 were:

Outcome Semaglutide 2.4 mg Semaglutide 1.0 mg Placebo
Average weight change −9.6% −7.0% −3.4%
Average HbA1c change −1.6 percentage points −1.5 percentage points −0.4 percentage points

Weight and blood sugar tell different parts of the story. The higher dose had a clearer advantage for weight; the two semaglutide groups had much closer blood-sugar changes.

Among people with week-68 weight measurements, about seven in ten in the 2.4 mg group lost at least 5%: 267/388, or 68.8%. That compared with 217/380, or 57.1%, with 1.0 mg and 107/376, or 28.5%, with placebo. These are observed counts, not percentages of everyone originally assigned.

The continued-treatment analysis estimated weight loss of 10.6%, 7.6% and 3.1%, respectively. Its assumptions explain why 10.6% sometimes appears beside the main 9.6% headline.

What side effects appeared?

About two in three people receiving 2.4 mg reported a digestive problem. Most were mild or moderate and temporary. Nausea was the most common.

The paper's safety table uses 403, 402 and 402 people, respectively: everyone who received at least one dose. One person in each assigned group never started.

Recorded problem Semaglutide 2.4 mg Semaglutide 1.0 mg Placebo
Any digestive problem 63.5% 57.5% 34.3%
Nausea 33.7% 32.1% 9.2%
Vomiting 21.8% 13.4% 2.7%
Serious health problem 9.9% 7.7% 9.2%
Stopped treatment because of a health problem 6.2% 5.0% 3.5%

A recorded health problem was not necessarily caused by the medicine. “Serious” includes events such as hospitalization; it is different from how unpleasant a symptom feels. One death was recorded in each of the three groups.

Severe low blood sugar or symptoms with a confirmed low reading occurred in 5.7%, 5.5% and 3.0%, respectively. That makes the other diabetes medicines part of the safety discussion.

What are the main cautions about the evidence?

The authors noted that excluding insulin users limited how widely the findings apply. It was not designed to settle whether treatment prevented heart attacks or strokes. Novo Nordisk funded the study and took part in its design and analysis.

A later pooled analysis by Wharton and colleagues, published in Diabetes, Obesity and Metabolism in 2022, examined STEP 1–3 together. Weight loss occurred with and without digestive side effects. These symptoms were most common during or shortly after dose increases. This reused trial data; it was not a new independent trial.

STEP 1 enrolled people without diabetes, so its larger weight-loss headline is not a direct comparison. The STEP 4 withdrawal trial asked about continuing versus stopping treatment after an initial treatment period. Neither design makes the three STEP 2 groups interchangeable.

What has changed in the safety picture?

As of November 12, 2026, the day this article was written, FDA's November 9 safety communication requires new eye-risk warnings for GLP-1 medicines, including semaglutide. These medicines mimic a gut hormone involved in blood sugar and appetite.

The rare condition, NAION, involves reduced blood flow to the nerve carrying sight and can cause permanent vision loss. FDA says to contact a health professional right away for sudden vision loss, blind spots, color changes or reduced sharpness. For patients who develop NAION, clinicians are told to consider stopping treatment unless the benefits clearly outweigh the risk.

What can a prescription discussion take from STEP 2?

The study helps separate weight goals, blood-sugar goals and treatment tolerance. The semaglutide and type 2 diabetes overview covers the broader clinical questions.

STEP 2 tested the manufacturer's study medicines, not a compounded preparation. Its averages cannot establish the results of a compounded version or predict one person's outcome.

Semaglutide prescribed through Promise is dispensed by a licensed U.S. compounding pharmacy as a compounded medication, which is different from an FDA-approved product: the formulation offered here is not FDA-approved. FDA explains that compounded medications do not undergo its review for safety, effectiveness or quality.

A licensed provider may still prescribe — that decision is between you and your doctor.

At Promise, a licensed provider reviews every request, and not everyone qualifies. The prescriber sets the actual dose and follow-up plan around the person's history and other medicines.