GLP-1 Receptor Agonists — Weight Loss Studies

The pivotal weight-loss trial program for GLP-1 receptor agonists is one of the most extensively studied pharmacotherapy programs in modern medicine. The headline numbers — 14.9% mean weight reduction with semaglutide 2.4 mg in STEP-1, 22.5% with tirzepatide 15 mg in SURMOUNT-1, 20% reduction in major adverse cardiovascular events in SELECT, 24% reduction in kidney-failure composite outcome in FLOW — have reshaped the standard of care for obesity and obesity-related disease. This page walks through the trial-by-trial evidence, the body-composition picture (not all the weight lost is fat), the rebound problem when therapy is discontinued, and the cardiovascular and renal outcome benefits that have moved GLP-1 agonists from "diabetes drug" to "metabolic and cardiovascular disease drug."

Table of Contents

  1. Trial Program Overview — STEP, SURMOUNT, SCALE, SELECT, FLOW
  2. STEP-1: Semaglutide 2.4 mg in Obesity Without Diabetes
  3. STEP-2 through STEP-8 Trials
  4. SCALE Obesity: Liraglutide 3.0 mg
  5. SURMOUNT-1: Tirzepatide and the 22.5% Number
  6. Body Composition: Fat vs Lean Mass
  7. Discontinuation and the Rebound Problem
  8. SELECT Cardiovascular Outcomes
  9. FLOW Renal Outcomes
  10. Key Research Papers
  11. Connections

Trial Program Overview — STEP, SURMOUNT, SCALE, SELECT, FLOW

The GLP-1 weight-loss trial program organizes around three sponsors and three trial families. Novo Nordisk runs the SCALE program (liraglutide 3.0 mg, branded Saxenda), the STEP program (semaglutide 2.4 mg, branded Wegovy), the SUSTAIN program (semaglutide for type 2 diabetes, branded Ozempic), the SELECT cardiovascular outcomes trial, and the FLOW renal outcomes trial. Eli Lilly runs the SURPASS program (tirzepatide for diabetes, branded Mounjaro) and the SURMOUNT program (tirzepatide for obesity, branded Zepbound). The two sponsors are now running head-to-head superiority trials (SURMOUNT-5 showed tirzepatide superior to semaglutide for weight loss in obesity without diabetes).

The program design is unusual for its size and the breadth of indications studied. The aggregate enrollment across STEP, SURMOUNT, SCALE, SELECT, FLOW, SUSTAIN, and SURPASS is over 50,000 patients followed for up to 5 years — an evidence base larger than for most cancer therapies. The trials cover obesity without diabetes (STEP-1, SURMOUNT-1, SCALE Obesity), obesity with diabetes (STEP-2, SURMOUNT-2), obesity with cardiovascular disease (SELECT), obesity with heart failure (STEP-HFpEF), obesity with knee osteoarthritis (STEP-9), obesity with NASH (SYNERGY-NASH, ESSENCE), and diabetes with chronic kidney disease (FLOW). Each indication has produced statistically and clinically significant outcome improvements.

Back to Table of Contents

STEP-1: Semaglutide 2.4 mg in Obesity Without Diabetes

STEP-1 (Wilding et al., NEJM 2021) was the registrational trial for semaglutide 2.4 mg weekly for obesity. The design enrolled 1,961 adults with a BMI of at least 30 (or at least 27 with weight-related comorbidity), without diabetes, and randomized 2:1 to semaglutide vs placebo on top of a 500 kcal/day deficit diet and 150 min/week of physical activity. Treatment was 68 weeks with a 16-week dose-escalation from 0.25 mg to 2.4 mg weekly.

Headline results at 68 weeks:

The 14.9% mean weight loss substantially exceeded all previous pharmacotherapy results (orlistat ~3%, phentermine-topiramate ~8%, naltrexone-bupropion ~5%) and approached bariatric-surgery territory. The trial established the benchmark against which all later obesity pharmacotherapy is compared.

Back to Table of Contents

STEP-2 through STEP-8 Trials

STEP-2 (Davies et al., Lancet 2021) replicated the STEP-1 design in 1,210 patients with both obesity and type 2 diabetes, comparing semaglutide 2.4 mg vs 1.0 mg vs placebo. Weight loss was 9.6% vs 7.0% vs 3.4% — the smaller absolute reduction in diabetic patients vs STEP-1 patients without diabetes is a consistent finding across the program. HbA1c reduction was −1.6% on the 2.4 mg dose.

STEP-3 (Wadden et al., JAMA 2021) tested whether intensive behavioral therapy (30 in-person sessions over 68 weeks plus a low-calorie meal-replacement plan) added to the semaglutide effect. Weight loss was 16.0% vs 5.7% on placebo — a slightly larger absolute effect than STEP-1, consistent with the principle that pharmacotherapy synergizes with structured lifestyle intervention.

STEP-4 (Rubino et al., JAMA 2021) was the seminal "what happens if you stop" trial. All 803 participants ran 20 weeks of open-label semaglutide titration to 2.4 mg, then were randomized to continue semaglutide vs switch to placebo for an additional 48 weeks. Patients who continued lost an additional 7.9%; those switched to placebo regained 6.9%. The difference in 48-week weight change between continuation and discontinuation was approximately 15 percentage points — the size of the original drug effect.

STEP-5 (Garvey et al., Nat Med 2022) extended treatment to 104 weeks. Mean weight loss was 15.2% on semaglutide vs 2.6% on placebo, demonstrating that the effect is largely sustained with continued dosing (no major attenuation beyond the first year).

STEP-HFpEF (Kosiborod et al., NEJM 2023) tested semaglutide in 529 patients with heart failure with preserved ejection fraction (HFpEF) and obesity. Semaglutide produced a 7.8-point greater improvement in the Kansas City Cardiomyopathy Questionnaire clinical summary score and 13.3% weight loss vs 2.6% on placebo — the first demonstration of GLP-1 agonist benefit in HFpEF specifically.

STEP-8 directly compared semaglutide 2.4 mg vs liraglutide 3.0 mg head-to-head, finding semaglutide superior at 68 weeks (15.8% vs 6.4% mean weight loss).

Back to Table of Contents

SCALE Obesity: Liraglutide 3.0 mg

Liraglutide 3.0 mg (Saxenda) was the first GLP-1 receptor agonist approved for chronic weight management, based on the SCALE program. The pivotal trial (Pi-Sunyer et al., NEJM 2015) enrolled 3,731 adults with obesity or overweight with comorbidity, without diabetes, and randomized 2:1 to liraglutide 3.0 mg daily vs placebo for 56 weeks on top of lifestyle modification.

Mean weight loss at 56 weeks was 8.0% on liraglutide vs 2.6% on placebo. The proportion losing at least 5% was 63.2% vs 27.1%, and the proportion losing at least 10% was 33.1% vs 10.6%. Adverse-event discontinuation was 9.9% vs 3.8%, again predominantly gastrointestinal. SCALE Diabetes (Davies et al., JAMA 2015) replicated the design in 846 patients with diabetes, producing 6.0% vs 2.0% weight loss. The smaller effect size vs semaglutide is consistent with liraglutide's shorter half-life and lower peak concentration.

Back to Table of Contents

SURMOUNT-1: Tirzepatide and the 22.5% Number

SURMOUNT-1 (Jastreboff et al., NEJM 2022) was the registrational obesity trial for tirzepatide. The design enrolled 2,539 adults with obesity or overweight with comorbidity, without diabetes, randomized 1:1:1:1 to tirzepatide 5 mg vs 10 mg vs 15 mg vs placebo weekly for 72 weeks on top of lifestyle modification.

Mean weight loss at 72 weeks was:

The 22.5% mean weight loss approached the 25–30% range typically seen with sleeve gastrectomy and represents the largest pharmacotherapy weight-loss effect ever documented. At the 15 mg dose, 91% of patients lost at least 5%, 78% at least 10%, 57% at least 20%, and 36% at least 25%. The trial set a new benchmark and supported FDA approval of tirzepatide (Zepbound) for obesity in late 2023.

SURMOUNT-2 (in patients with type 2 diabetes) showed 12.8% vs 3.2% weight loss; SURMOUNT-3 (with intensive lifestyle lead-in) showed an additional 18.4% loss after the lead-in vs continued placebo; SURMOUNT-4 (Aronne et al., JAMA 2024) was the tirzepatide withdrawal trial, showing 14% weight regain after switching from tirzepatide to placebo over 52 weeks.

Back to Table of Contents

Body Composition: Fat vs Lean Mass

The 15–22% total-weight loss in the headline trials is not entirely fat mass. DEXA substudies in STEP-1 and SURMOUNT-1 documented that approximately 25–40% of the lost weight is lean mass (muscle, organ, water, glycogen-bound water). For a patient losing 30 kg total, this can mean 8–12 kg of lean mass loss — equivalent to the entire skeletal muscle of one leg. The remaining 60–75% of lost weight is fat mass, which is the desirable component.

The lean-mass loss is a normal feature of any large weight loss — bariatric surgery patients show similar proportions, and even unmedicated dieting produces about 25% lean-mass loss. The concern with GLP-1 agonists is that the magnitude of total loss makes the absolute lean-mass loss large in older patients, who start with less muscle reserve and have a harder time rebuilding it. The clinical guidance now emphasizes resistance training and adequate protein intake (1.2–1.6 g/kg/day) throughout the weight-loss phase to minimize lean-mass loss.

Several adjunctive strategies are being studied. Bimagrumab (an activin/myostatin receptor blocker) preserved lean mass and accelerated fat loss in a small phase-2 trial in combination with semaglutide. Resistance training with progressive overload is the cheapest and best-validated intervention. Protein supplementation alone (whey or casein) does not appear sufficient without the training stimulus.

Back to Table of Contents

Discontinuation and the Rebound Problem

The single most consequential finding from the trial program for long-term patient counseling is the rebound phenomenon documented in STEP-4 (Rubino, JAMA 2021), STEP-1 extension (Wilding, Diabetes Obes Metab 2022), and SURMOUNT-4 (Aronne, JAMA 2024). All three trials show that stopping the medication results in substantial weight regain over the subsequent 12 months — typically two thirds of the weight lost.

The mechanism appears to be a combination of restored appetite (loss of the central GLP-1 receptor stimulation that suppressed it), restored gastric emptying (loss of the satiety signal from delayed emptying), and the body's defended-weight set-point biology — metabolic adaptation reduces resting energy expenditure during weight loss and persists for years, creating a powerful counter-regulatory drive back toward the original weight. This biology operates after bariatric surgery as well but is partially defended by the anatomic restriction; with pharmacotherapy, removing the drug removes the only counter-force.

The practical implication is that GLP-1 receptor agonist therapy for obesity is, like therapy for hypertension or dyslipidemia, a chronic indefinite treatment. The framing "take the drug for a year and stop" does not match the biology — it produces a year of weight loss followed by a year of weight regain. Patients and physicians both benefit from understanding this at the outset.

Back to Table of Contents

SELECT Cardiovascular Outcomes

SELECT (Lincoff et al., NEJM 2023) was the landmark cardiovascular outcomes trial for semaglutide in obesity without diabetes. The design enrolled 17,604 patients with established cardiovascular disease and a BMI of at least 27, randomized 1:1 to semaglutide 2.4 mg weekly vs placebo. Median follow-up was 39.8 months.

The primary composite endpoint (cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke) occurred in 6.5% of the semaglutide group vs 8.0% of placebo (HR 0.80, 95% CI 0.72–0.90, p<0.001) — a 20% relative risk reduction over 3.3 years. The absolute risk reduction was 1.5 percentage points, yielding a number-needed-to-treat of 67. Mean weight loss was 9.4% on semaglutide vs 0.9% on placebo at 104 weeks. Notably, the cardiovascular benefit was independent of baseline glycemic status (the trial enrolled patients without diabetes) and appeared to emerge before substantial weight loss had accumulated, suggesting mechanisms beyond weight loss alone.

The SELECT result moved semaglutide from a metabolic drug to a cardiovascular drug. The FDA expanded the Wegovy label in March 2024 to include reduction of cardiovascular events in adults with established CVD and obesity or overweight. This established the precedent for insurance coverage of obesity pharmacotherapy on cardiovascular grounds.

Back to Table of Contents

FLOW Renal Outcomes

FLOW (Perkovic et al., NEJM 2024) was the kidney-outcomes trial for semaglutide in patients with type 2 diabetes and chronic kidney disease (eGFR 25–75 mL/min/1.73 m² with albuminuria). The trial enrolled 3,533 patients and was stopped early in late 2023 for efficacy.

The primary composite endpoint — kidney failure (sustained eGFR <15, dialysis, or transplantation), at least 50% sustained reduction in eGFR, or kidney-related or cardiovascular death — occurred 24% less often on semaglutide (HR 0.76, 95% CI 0.66–0.88, p=0.0003). Mean eGFR slope was −2.19 vs −3.36 mL/min/1.73 m²/year — a slowing of decline by 35%. The benefit was on top of optimized ACE-inhibitor / ARB therapy in the great majority of patients, and was statistically independent of SGLT2 inhibitor co-treatment.

FLOW makes GLP-1 receptor agonists the third drug class with documented benefit on hard kidney outcomes in diabetic CKD (after ACE-inhibitors/ARBs and SGLT2 inhibitors), and supports a quadruple-therapy paradigm for high-risk CKD patients with diabetes.

Back to Table of Contents

Key Research Papers

  1. Wilding JPH et al., STEP-1 semaglutide 2.4 mg in obesity (NEJM 2021) — PubMed: Wilding STEP-1
  2. Jastreboff AM et al., SURMOUNT-1 tirzepatide in obesity (NEJM 2022) — PubMed: Jastreboff SURMOUNT-1
  3. Pi-Sunyer X et al., SCALE Obesity liraglutide 3.0 mg (NEJM 2015) — PubMed: Pi-Sunyer SCALE
  4. Wadden TA et al., STEP-3 semaglutide + behavioral therapy (JAMA 2021) — PubMed: Wadden STEP-3
  5. Rubino D et al., STEP-4 semaglutide withdrawal (JAMA 2021) — PubMed: Rubino STEP-4
  6. Garvey WT et al., STEP-5 104-week semaglutide (Nat Med 2022) — PubMed: Garvey STEP-5
  7. Davies M et al., STEP-2 semaglutide in T2D (Lancet 2021) — PubMed: Davies STEP-2
  8. Aronne LJ et al., SURMOUNT-4 tirzepatide withdrawal (JAMA 2024) — PubMed: Aronne SURMOUNT-4
  9. Lincoff AM et al., SELECT semaglutide CV outcomes (NEJM 2023) — PubMed: Lincoff SELECT
  10. Perkovic V et al., FLOW semaglutide renal outcomes (NEJM 2024) — PubMed: Perkovic FLOW
  11. Kosiborod MN et al., STEP-HFpEF semaglutide in HFpEF (NEJM 2023) — PubMed: Kosiborod STEP-HFpEF
  12. Rubino DM et al., STEP-8 semaglutide vs liraglutide (JAMA 2022) — PubMed: Rubino STEP-8
  13. Wilding JPH et al., STEP-1 extension weight regain (Diabetes Obes Metab 2022) — PubMed: STEP-1 extension

Back to Table of Contents

Connections

Back to Table of Contents