Half the people who start a GLP-1 drug report a gut symptom within five years. When doctors actually scanned the stomachs of the ones who complained, two out of three were emptying food at a normal speed.
That gap, between what patients feel and what the stomach is measurably doing, is the finding in a Mayo Clinic study of 78,277 GLP-1 users published this month in Neurogastroenterology and Motility ↗. It is the largest real-world look yet at the side effect everyone who has taken semaglutide or a cousin drug already knows about: the nausea, the early fullness, the churn.
What the numbers say
The Mayo team pulled records from the Mayo Clinic Platform database and tracked how long patients went before their first gastrointestinal symptom event, a category the authors call GISE. At the start, 87 percent of the cohort had no gut symptoms on record. By 60 months, 53 percent had logged at least one. The most common were abdominal pain, in 28 percent, and nausea, in 25 percent. Women were about 40 percent more likely to hit a symptom than men (hazard ratio 1.4), the kind of sex gap that keeps showing up across GLP-1 tolerability data.
What predicted trouble was mostly a history of trouble. Patients who already had diarrhea before starting were more than twice as likely to log a symptom (hazard ratio 2.2), and prior constipation, abdominal pain, and nausea each raised the odds too (2.0, 1.7, and 1.6). Taking metoclopramide, a drug given to speed up a sluggish gut, at baseline carried its own elevated risk (1.6), which reads less like cause and more like a marker for people whose stomachs were already complaining. Even with all of that, the prediction model was weak. Its discrimination score landed at 0.64, barely better than a coin-weighted guess. The honest read is that doctors still cannot tell you in advance whether your stomach will tolerate these drugs.
The part that cuts against the story
Here is the sentence worth arguing with. Among 696 patients who were symptomatic enough that their doctors sent them for a gastric emptying scan, only 34.6 percent, 241 people, actually showed delayed emptying. The other two-thirds had stomachs clearing food on schedule while still feeling sick.
That matters because the popular explanation for GLP-1 gut trouble is that the drugs slow the stomach to a crawl, the mechanism that also drives the weight loss and that sits behind anesthesiologists' worries about food left in the stomach during surgery. This platform has quantified that delay before ↗: controlled studies put it around 74 minutes on average. The new data does not overturn that. It adds a wrinkle. In the patients sick enough to get tested, slow emptying was the exception, not the rule. The symptoms are real, but a measurable stomach-motor problem explains only a minority of them. Constipation was the one symptom more common in the delayed-emptying group, which fits, since a slow stomach often travels with a slow gut below it.
Why it is worth your attention
The selection here is the obvious caveat. Only symptomatic patients got referred for a scan, so the 34.6 percent is not the rate across everyone on the drug, just across the complainers who reached a specialist. The study is retrospective, built from what clinicians happened to chart, and a database cannot see dose, adherence, or how fast someone titrated up. None of that softens the central point. For a class of drugs now taken by tens of millions, the most common reason people quit is a symptom we cannot predict well and often cannot pin on the one mechanism everyone assumes.
For the broader pipeline, that is a design target, not just a footnote. Semaglutide ↗ and the dual GLP-1 and GIP agonist tirzepatide ↗ both act on the GLP-1 receptor ↗ that sets gut motility, and the next wave of candidates is being pitched partly on gentler stomachs. If half of real-world users feel something within five years and two-thirds of the sickest are not emptying slowly, the companies chasing a cleaner tolerability profile are aiming at a problem that is only partly about speed.