GLP-1 medications

Large placebo-controlled trials in adults without diabetes found substantial weight loss alongside lifestyle counseling. Semaglutide 2.4 mg weekly (STEP 1, 1,961 adults, 68 weeks): 14.9% vs 2.4% on placebo, and half of people lost 15% or more. Tirzepatide weekly (SURMOUNT-1, 2,539 adults, 72 weeks): 15.0%, 19.5% and 20.9% at 5, 10 and 15 mg vs 3.1%. Liraglutide 3.0 mg daily (SCALE Obesity and Prediabetes, 3,731 adults, 56 weeks): 8.4 kg vs 2.8 kg. Oral semaglutide 25 mg daily (OASIS 4, 307 adults, 64 weeks): 13.6% vs 2.2%. Orforglipron daily (ATTAIN-1, 3,127 adults, 72 weeks, investigational formulation at 6 to 36 mg): 7.5% to 11.2% depending on dose vs 2.1%. These figures come from separate trials and cannot be compared directly. A 2026 network meta-analysis of 262 trials rated the one-year weight-loss evidence for tirzepatide, semaglutide and orforglipron as moderate to high certainty, and found that larger benefits generally came with more side effects and more people stopping treatment.

Three trials tested what happens when the drug stops. In the STEP 1 extension (327 adults), people who had lost 17.3% on semaglutide regained 11.6 percentage points within a year of stopping it and lifestyle support, about two-thirds of the loss, and most heart and metabolic improvements moved back toward baseline. In STEP 4, after 20 weeks on semaglutide (10.6% loss), 803 adults were randomized: those who continued lost a further 7.9% over 48 weeks, while those switched to placebo regained 6.9%. In SURMOUNT-4, after 36 weeks on tirzepatide (20.9% loss), 670 adults were randomized: continuing gave a further 5.5% loss, while switching to placebo led to a 14.0% weight gain over 52 weeks. The weight-management labels describe use to "maintain weight reduction long term".

A 2017 network meta-analysis of 34 randomized trials (14,464 people) found that every GLP-1 medication studied lowered A1c compared with placebo over about 24 to 32 weeks, by 0.55 to 1.21 percentage points depending on the drug; the long-acting products (dulaglutide, liraglutide, weekly exenatide) lowered it more than twice-daily exenatide. Semaglutide was not yet included, and the drugs raised the risk of low blood sugar and stomach side effects compared with placebo. In SURPASS-2 (1,879 adults, average starting A1c 8.28%), A1c fell 1.86 points with semaglutide 1 mg weekly and 2.01 to 2.30 points with tirzepatide 5 to 15 mg over 40 weeks; blood sugar below 54 mg/dL occurred in 0.2% to 1.7% on tirzepatide and 0.4% on semaglutide.

Several large placebo-controlled trials in people at high risk found fewer major cardiovascular events (heart death, nonfatal heart attack or nonfatal stroke). In type 2 diabetes: liraglutide in LEADER (9,340 people, median 3.8 years), 13.0% vs 14.9% (hazard ratio 0.87); weekly semaglutide in SUSTAIN-6 (3,297 people, 2 years), 6.6% vs 8.9% (0.74); dulaglutide in REWIND (9,901 people, median 5.4 years), 12.0% vs 13.4% (0.88); oral semaglutide in SOUL (9,650 people, about 4 years), 12.0% vs 13.8% (0.86). Without diabetes: semaglutide 2.4 mg in SELECT (17,604 people aged 45 or older with heart disease and a BMI of 27 or more, about 3.3 years), 6.5% vs 8.0% (0.80). Tirzepatide was compared with dulaglutide rather than placebo in SURPASS-CVOT (13,165 people with diabetes and heart disease): 12.2% vs 13.1%, which showed it was not worse but did not prove it better. These benefits apply to people with existing heart disease or high risk, the groups studied.

Observational studies had suggested less dementia in people taking GLP-1 drugs. The evoke and evoke+ trials tested this directly in 3,808 people aged 55 to 85 with amyloid-confirmed mild cognitive impairment or mild dementia due to Alzheimer's disease, randomized to oral semaglutide up to 14 mg daily or placebo. After 104 weeks, decline on the Clinical Dementia Rating scale was the same in both groups in both trials (differences of -0.08 and 0.10 points, not significant), and both trials were stopped.

SURMOUNT-5 is the only trial that directly compared the two weekly injections for weight. It randomized 751 adults with obesity but without diabetes to the highest tolerated dose of tirzepatide (10 or 15 mg) or semaglutide (1.7 or 2.4 mg) for 72 weeks. Weight fell 20.2% with tirzepatide and 13.7% with semaglutide, and waist size fell 18.4 cm vs 13.0 cm. Stomach and bowel side effects were the most common in both groups. The trial was open label (participants knew which drug they got) and was funded by the maker of tirzepatide. In people with type 2 diabetes, the SURPASS-2 trial found tirzepatide also lowered A1c and weight more than semaglutide 1 mg.

EXSCEL randomized 14,752 people with type 2 diabetes, 73.1% with prior cardiovascular disease, to once-weekly extended-release exenatide 2 mg or placebo for a median of 3.2 years. Major cardiovascular events occurred in 11.4% vs 12.2% (hazard ratio 0.91, 95% CI 0.83 to 1.00); the drug was safe for the heart but not significantly better than placebo. FDA now lists the once-weekly exenatide brand (Bydureon BCise) as discontinued.

FLOW randomized 3,533 adults with type 2 diabetes and chronic kidney disease to semaglutide 1.0 mg weekly or placebo. The trial was stopped early for benefit after a median of 3.4 years. Major kidney events (kidney failure, a 50% fall in kidney function, or death from kidney or heart causes) were 24% less likely with semaglutide (hazard ratio 0.76), the yearly decline in kidney function was slower, and death from any cause was 20% lower. This trial supports the Ozempic indication for kidney disease in type 2 diabetes. A 2026 network meta-analysis of obesity drugs found no drug that convincingly reduced kidney failure in people with obesity, so the evidence is specific to diabetic kidney disease.

SURMOUNT-OSA was two 52-week trials of tirzepatide (10 or 15 mg weekly) vs placebo in adults with moderate to severe sleep apnea and obesity (average BMI about 39). Breathing interruptions per hour of sleep (the apnea-hypopnea index, about 50 at the start) fell by 25.3 vs 5.3 events per hour in people not using a breathing machine, and by 29.3 vs 5.5 in people using one. Oxygen drops, blood pressure and sleep-related symptoms also improved. This led to the Zepbound sleep apnea approval on December 20, 2024.

In STEP-HFpEF, 529 people with this type of heart failure and a BMI of 30 or more took semaglutide 2.4 mg or placebo for 52 weeks. Symptom and physical-limitation scores (KCCQ, 0 to 100) improved by 16.6 vs 8.7 points, and 6-minute walk distance improved by 20.3 m more than placebo. In SUMMIT, 731 similar patients took tirzepatide or placebo for a median of 104 weeks: heart death or worsening heart failure occurred in 9.9% vs 15.3% (hazard ratio 0.62), and symptom scores improved 6.9 points more. In SUMMIT, heart deaths alone were few and not lower (8 vs 5 people). Neither Wegovy nor Zepbound has an FDA indication for heart failure, and the Wegovy label states that an effect on heart failure has not been established.

In the ongoing ESSENCE trial, 1,197 adults with biopsy-proven MASH and moderate to advanced scarring (fibrosis stage 2 or 3) were assigned to semaglutide 2.4 mg weekly or placebo. In the first 800 patients at 72 weeks, liver inflammation resolved without worse scarring in 62.9% vs 34.3%, and scarring improved without worse inflammation in 36.8% vs 22.4%. These are biopsy results, not proof of fewer cases of cirrhosis, liver failure or death; FDA granted Wegovy an accelerated approval for MASH on August 15, 2025, and continued approval may depend on a confirmatory trial showing clinical benefit.