The long trials ran for years, not decades. The longest we list followed people for a median 5.4 years (REWIND, dulaglutide 1.5 mg). Semaglutide at the weight-loss dose has been followed for a mean 39.8 months (SELECT) and tirzepatide for 176 weeks against placebo. In SURMOUNT-1 the most common side effects were stomach ones, mostly in the first 20 weeks. In SELECT, side effects made 16.6% on semaglutide stop for good, against 8.2% on placebo. What happens over decades is not known, and no long trial tested a compounded version.
Figures read in each paper's PubMed abstract, Nov 6, 2026.
Each compared the medicine with placebo. A heart or kidney trial enrolls people who are sicker than most buyers, so its side effect rates are not yours.
REWIND: median 5.4 years
Dulaglutide 1.5 mg, 9,901 people. Type 2 diabetes.
A gastrointestinal side effect was reported by 47.4% on dulaglutide and 34.1% on placebo. Deaths from any cause did not differ.
SOUL: median 49.5 months
Oral semaglutide, up to 14 mg, 9,650 people. Type 2 diabetes with heart or kidney disease.
Serious adverse events: 47.9% on oral semaglutide, 50.3% on placebo.
LEADER: median 3.8 years
Liraglutide, 9,340 people. Type 2 diabetes with high heart risk.
Stomach problems were the most common reason for stopping. Pancreatitis was not significantly different from placebo.
FLOW: median 3.4 years
Semaglutide, 3,533 people. Type 2 diabetes with chronic kidney disease.
Serious adverse events: 49.6% on semaglutide, 53.8% on placebo.
SURMOUNT-1, three-year part: 176 weeks
Tirzepatide 5, 10 or 15 mg, 1,032 people. Obesity with prediabetes.
Stomach side effects were most common, mostly mild to moderate and mostly in the first 20 weeks. The authors report no new safety signals.
SELECT: mean 39.8 months
Semaglutide 2.4 mg, 17,604 people. Heart disease and overweight or obesity, no diabetes.
Side effects made 16.6% on semaglutide and 8.2% on placebo stop for good.
STEP 5: 104 weeks
Semaglutide 2.4 mg, 304 people. Overweight or obesity, no diabetes.
A gastrointestinal side effect was reported by 82.2% on semaglutide and 53.9% on placebo, mostly mild to moderate.
SURPASS-CVOT compared tirzepatide with dulaglutide, not placebo, in 13,299 people with type 2 diabetes and heart disease. It ran from May 2020 to June 2025. Its paper says: “The incidence of adverse events appeared to be similar in the two groups, although more gastrointestinal adverse events were observed in the tirzepatide group.”
Not a trial: it compares people who chose or were prescribed different medicines, so it can show a link and not a cause.
Researchers followed 215,970 US veterans with diabetes who started a GLP-1, across 175 health outcomes. Against usual care, they found more gastrointestinal disorders, hypotension, syncope, arthritic disorders, nephrolithiasis, interstitial nephritis and drug-induced pancreatitis, and less substance use and psychotic disorders, seizures, neurocognitive disorders (including Alzheimer's disease and dementia), coagulation disorders, cardiometabolic disorders, infectious illnesses and several respiratory conditions. Nature Medicine, 2025
Where studies disagree, both sides are here, each with what kind of study it is.
The long trials ran for years, not decades. The longest we list followed people for a median 5.4 years (REWIND, dulaglutide 1.5 mg). Semaglutide at the weight-loss dose has been followed for a mean 39.8 months (SELECT) and tirzepatide for 176 weeks against placebo. In SURMOUNT-1 the most common side effects were stomach ones, mostly in the first 20 weeks. In SELECT, side effects made 16.6% on semaglutide stop for good, against 8.2% on placebo. What happens over decades is not known, and no long trial tested a compounded version. One observational study mapped this: among 215,970 US veterans with diabetes who started one, across 175 outcomes, it found more gastrointestinal disorders, hypotension, syncope, arthritic disorders, nephrolithiasis, interstitial nephritis and drug-induced pancreatitis than usual care. That shows an association, not a cause, and these people are not everyone who takes one. Nature Medicine, 2025
For semaglutide, the longest study we list is SOUL: 9,650 people, median 49.5 months, in type 2 diabetes with heart or kidney disease. Serious adverse events: 47.9% on oral semaglutide, 50.3% on placebo. SELECT, in people without diabetes at the 2.4 mg dose of Wegovy, is the closest to someone taking it for weight. SOUL on PubMed
For tirzepatide, the longest study we list is SURMOUNT-1, three-year part: 1,032 people, 176 weeks, in obesity with prediabetes. Stomach side effects were most common, mostly mild to moderate and mostly in the first 20 weeks. The authors report no new safety signals. A larger trial against another GLP-1, SURPASS-CVOT with 13,299 people, ran from May 2020 to June 2025: "The incidence of adverse events appeared to be similar in the two groups, although more gastrointestinal adverse events were observed in the tirzepatide group." SURPASS-CVOT on PubMed
The labels for weight loss describe long-term use and give no end date. Wegovy: "to reduce excess body weight and maintain weight reduction long term" Zepbound: "to reduce excess body weight and maintain weight reduction long term" Foundayo: "to reduce excess body weight and maintain weight reduction long term" Trials show what happened over the years they ran, listed above, and people who stopped mostly regained weight. Whether to stay on one is for you and your prescriber. What happens when you stop
A nickname for a thinner, older-looking face after weight loss, not a term in any FDA label we read. It is a nickname, for a thinner face after weight loss. In 20 people on GLP-1s who had head scans before and after, the middle of the face lost a median 9.0% of its volume, about 7% for every 10 kg lost. A small look-back study, with no comparison group. Otolaryngol Head Neck Surg, 2025
Nobody has measured them. The trials here tested the approved products, and compounded drugs are not FDA-approved: the FDA does not review them for safety, effectiveness or quality before they are marketed. A compounded version has the same medicine named on its label, but its long-term record cannot be read across with certainty. FDA: concerns with unapproved GLP-1 drugs used for weight loss