Semaglutide — Clinical Evidence
All published trials referenced on this site. Click any source link to read the original journal article.
Last reviewed: March 2026
Randomised controlled trials
In the following trials, participants were randomly assigned to Semaglutide or a dummy pill (placebo). All weight loss figures are average body weight reductions from baseline.
| Trial | Key result | Source |
|---|---|---|
| [1]STEP 1 | 14.9% mean weight loss. 86% achieved ≥5% loss; 69% achieved ≥10%; 50% achieved ≥15%. Nausea 44% vs. 16% placebo; diarrhoea 30% vs. 16%; vomiting 24% vs. 6%; constipation 24% vs. 11%. | New England Journal of Medicine 2021 ↗ |
| [2]STEP 2 | 9.6% mean weight loss — lower than STEP 1 because type 2 diabetes reduces GLP-1 response. Significant HbA1c improvements. | New England Journal of Medicine 2021 ↗ |
| [3]STEP 5 | 15.2% sustained mean weight loss at 2 years — showing durable effect with continued use. | Nature Medicine 2022 ↗ |
| [4]OASIS 4 (oral semaglutide 25mg) | 13.6% average weight loss (all participants, including those who reduced dose or stopped). 16.6% average weight loss with full adherence. 76.3% achieved ≥5% weight loss vs. 31.3% placebo. About 1 in 3 adherent participants achieved ≥20% weight loss. | New England Journal of Medicine 2024 ↗ |
| [5]SELECT (cardiovascular outcomes) | 20% reduction in MACE (non-fatal heart attack, non-fatal stroke, or cardiovascular death) vs. placebo — the first weight loss drug to demonstrate CV event reduction. | New England Journal of Medicine 2023 ↗ |
Reported side effects
Frequencies from STEP 1. View source ↗
| Side effect | Frequency in trial |
|---|---|
| Nausea | 44% |
| Diarrhoea | 30% |
| Vomiting | 24% |
| Constipation | 24% |
Body composition data
DXA substudy data from STEP 1 showed semaglutide led to proportionally greater reduction in fat mass than lean mass — consistent with findings from other GLP-1 drugs. Lean mass loss was proportional to the amount of weight lost overall.
Cardiovascular & metabolic data
The SELECT trial (n=17,604) demonstrated a 20% reduction in major adverse cardiovascular events (MACE — heart attack, stroke, or cardiovascular death) vs. placebo in adults with established cardiovascular disease and obesity, without diabetes. Also showed fewer heart failure events in STEP-HFpEF.