No obesity drug has ever been tested this heavily. Between 2021 and 2024 the STEP program ran a coordinated series of phase 3 trials that worked through the obvious questions one at a time: does it work, does it work in diabetes, does it beat lifestyle therapy, does the weight come back, does it last two years, does it beat the older drug, does it work in teenagers, does it help a failing heart. Then SELECT asked whether it stops people having heart attacks, the question that decides whether a weight drug is a medicine. It does. What follows is those trials in the order they answered something, with the number each one is remembered for.
Ranked papers
#1STEP-1
Wilding JPH, Batterham RL, Calanna S, et al. · N Engl J Med · 2021
What it measured: Body weight change at 68 weeks
The trial everything else is measured against. 1,961 adults carrying obesity without diabetes, given weekly semaglutide 2.4 mg against placebo across 68 weeks, and the result was −14.9% against −2.4%. A 15% loss or more was reached by just over half the treated group and by one in twenty on placebo. This is where the Wegovy 2.4 mg dose comes from, and it reset what counts as a good result in obesity medicine.
PMID 33567185 ↗NCT03548935 ↗
#2STEP-2
Davies M, Færch L, Jeppesen OK, et al. · Lancet · 2021
What it measured: Body weight change at 68 weeks
STEP-2 ran the same design in 1,210 people who also had type 2 diabetes, a group that consistently loses less weight on these drugs for reasons that are still argued about. Semaglutide delivered −9.6% against −3.4%, roughly two-thirds of what STEP-1 achieved, alongside a 1.6-point fall in A1C. Diabetes blunts the effect without abolishing it, which is why the same person may be a candidate for the weight-labeled product rather than the diabetes-labeled one.
PMID 33667417 ↗NCT03552757 ↗
#3STEP-3
Wadden TA, Bailey TS, Billings LK, et al. · JAMA · 2021
What it measured: Body weight change at 68 weeks
The question here was whether the drug and serious behavioral therapy overlap or add. Everyone got the intensive program (30 counseling visits, with a low-calorie diet for the first eight weeks), and two-thirds were randomized to semaglutide. The drug arm lost 16.0% and the therapy-only arm 5.7%, so the two add. That finding matters in the insurance argument, because prior-authorization criteria routinely demand a concurrent lifestyle program, and this is the evidence that the requirement is not merely bureaucratic.
PMID 33625476 ↗NCT03611582 ↗
#4STEP-4
Rubino D, Abrahamsson N, Davies M, et al. · JAMA · 2021
What it measured: Weight lost after the first 20 weeks, through week 68
The withdrawal trial, and the most underrated one on this list. Participants first spent 20 weeks getting to the full dose, losing about 10.6% along the way. Then two-thirds kept going and one-third switched to placebo. The continuers lost a further 7.9%, while the switchers regained 6.9%. Most of the lost weight came back within 48 weeks of stopping a drug that had been working. That is the finding that makes obesity a chronic condition rather than a course of treatment.
PMID 33755728 ↗NCT03548987 ↗
#5STEP-5
Garvey WT, Batterham RL, Bhatta M, et al. · Nat Med · 2022
What it measured: Body weight change at 104 weeks
STEP-1 extended to two years, answering the durability objection raised the moment the first result appeared. At 104 weeks the treated group held −15.2% against −2.6%, and half of those who completed were still down at least 15%. No randomized trial of this drug has run to five years, so when someone asks what happens after five years, the answer is that nobody has run that trial.
PMID 36216945 ↗NCT03693430 ↗
#6STEP-8
Rubino DM, Greenway FL, Khalid U, et al. · JAMA · 2022
What it measured: Body weight change at 68 weeks
The first head-to-head between two approved GLP-1 obesity drugs, and it was decisive. Weekly semaglutide produced −15.8% against daily liraglutide's −6.4% over 68 weeks, a gap of more than nine percentage points. Roughly seven in ten reached at least 10% loss on semaglutide against one in four on liraglutide, and more than twice as many people abandoned liraglutide. It effectively ended the case for starting anyone on the older drug.
PMID 35015037 ↗NCT04074161 ↗
#7STEP-TEENS
Weghuber D, Barrett T, Barrientos-Pérez M, et al. · N Engl J Med · 2022
What it measured: BMI change at 68 weeks
201 adolescents aged 12 to 17, and the effect was, if anything, larger than in adults: −16.1% BMI against a placebo group that drifted slightly upward. This is the trial behind the FDA's December 2022 decision to extend Wegovy to children aged 12 and over, making it the first weekly GLP-1 with that indication. Daily liraglutide (Saxenda) had been approved for the same age group in December 2020.
PMID 36322838 ↗NCT04102189 ↗
#8STEP-HFpEF
Kosiborod MN, Abildstrøm SZ, Borlaug BA, et al. · N Engl J Med · 2023
What it measured: How patients rated their own heart-failure symptoms, plus weight
529 participants, every one with obesity and the preserved-ejection-fraction form of heart failure (HFpEF), a form of the condition that had no therapy approved to target it. Semaglutide improved symptom scores by 16.6 points against 8.7 on placebo (a 7.8-point difference), with weight loss of 13.3% against 2.6%. The significance is conceptual as much as clinical: it is the first strong evidence that these drugs treat a disease of the heart rather than merely the weight sitting on top of it.
PMID 37622681 ↗NCT04788511 ↗
#9SELECT
Lincoff AM, Brown-Frandsen K, Colhoun HM, et al. · N Engl J Med · 2023
What it measured: First cardiac death, heart attack or stroke
The trial that turned a weight drug into cardiovascular medicine. 17,604 adults without diabetes who had established heart disease along with overweight or obesity, followed for years. Major cardiovascular events occurred in 6.5% on semaglutide against 8.0% on placebo (hazard ratio 0.80), a fifth fewer heart attacks, strokes and cardiovascular deaths. This produced the March 2024 cardiovascular indication, and it is the reason insurers and Medicare can pay for the drug on grounds that have nothing to do with the scale. Note the other number too: 16.6% stopped it for side effects, against 8.2% on placebo.
PMID 37952131 ↗NCT03574597 ↗
#10OASIS-1
Knop FK, Aroda VR, do Vale RD, et al. · Lancet · 2023
What it measured: Body weight change at 68 weeks
The high-dose oral trial: 50 mg of semaglutide swallowed daily by 667 adults with obesity, producing −15.1% against −2.4%, a result that stands comfortably alongside the injection. Wegovy tablets have since been approved, but the approved tablet reaches a 25 mg maintenance dose, not the 50 mg tested here. The oral market it was built to enter now also contains orforglipron, a different molecule entirely.
PMID 37385278 ↗NCT05035095 ↗
About this list
Obesity medicine throws up the same handful of questions over and over, and we answer them with ranked lists anchored to the papers themselves. On 2026-10-30 every citation below was put back through PubMed. Where a trial has a registry entry, the card carries that link too, alongside the PubMed one.
Browse our full index of research lists or our long-form research articles.
See also: our semaglutide overview page.