How Often Should You Poop on a GLP-1? What the Labels Show

Type “how often should you poop on” into Google and let it finish the sentence. Seven of the ten suggestions that come back are the names of GLP-1 medications. It is one of the most-asked gut questions of the moment, and the pages that answer it are almost entirely companies that would like to sell you a prescription.

So here is the version assembled from the drug labels instead, which are public, free, and apparently unread. It contains one answer nobody wants (there is no GLP-1-specific number), two findings that run backwards from the popular story, and a correction to the statistic you have probably already seen.

The baselineThere is no special number, and that is the answer

Start with what normal means for anyone, on any medication or none. The most useful modern figure comes from an analysis of NHANES data covering 4,775 US adults: 95.9 percent reported somewhere between three and twenty-one bowel movements a week. Three a day to three a week — the range clinicians have used for decades, confirmed on a nationally representative sample.

That range does not narrow because you started a medication. There is no published GLP-1-specific frequency, no study establishing what is typical on semaglutide, and no threshold that means anything on its own. What changes on these drugs is not the definition of normal but the odds of drifting toward one end of it — and the labels put real numbers on those odds. Our frequency checker lays out the population data in more detail, and the wider question has its own answer.

The labelsFour drugs, both directions, in public

Every GLP-1 label records constipation and diarrhoea, because the trials recorded both. Here they are together, with the placebo arm alongside each one — which is the number that turns a scary percentage into a meaningful one.

Two things in that table are worth sitting with, and neither appears in the coverage.

The first: on Zepbound, constipation goes down as the dose goes up — 17 percent at 5 mg, 14 at 10, 11 at 15 — while diarrhoea climbs from 19 to 23 percent over the same range. The gut is not simply getting more of one effect at higher doses. It is trading one for the other, and anyone who assumed a bigger dose meant more constipation had it backwards.

The second: the pill is the hardest on the gut, not the easiest. Foundayo — orforglipron, approved on 1 April 2026 as the first GLP-1 tablet for weight loss that can be taken without food or water restrictions — carries the highest constipation rate of the four at 20 to 27 percent depending on dose. Across its two pooled trials, gastrointestinal reactions of some kind occurred in 60 to 69 percent of people taking it against 37 percent on placebo, and 8 percent stopped treatment because of side effects against 3 percent on placebo. The convenience story and the tolerability story point in opposite directions, which is worth knowing before the pill is framed as the gentle option. We covered its side-effect profile in full when it launched.

The mechanismEveryone blames the colon. The evidence is about the stomach.

Here is the correction that matters most, and it is not a small one.

The universal explanation for GLP-1 constipation is that these drugs slow the gut down, so things move through the colon more slowly. The first half is measured. The second half is not.

What has been demonstrated, repeatedly, is delayed gastric emptying — how fast the stomach hands food onward. In a scintigraphy study of 20 women on semaglutide 1.0 mg, 37 percent of a solid meal was still sitting in the stomach four hours in, against zero percent on placebo, with half-emptying time running 171 minutes versus 118. That is a striking result, in a small, single-sex, PCOS-specific sample.

Now the part that gets left out. That delay fades. In a 72-person study at the 2.4 mg weight-management dose, the effect on gastric emptying was not detectable at all after 20 weeks — and the Ozempic label states plainly that “no apparent effect on the rate of gastric emptying was observed with semaglutide 2.4 mg.” Tirzepatide’s label says the delay is largest after the first dose and diminishes over time. The idea that these drugs permanently freeze your digestion is accurate for the early weeks and increasingly wrong after that.

And the colon? Nobody has looked. Across semaglutide, tirzepatide and orforglipron there is no published human study of colonic transit or bowel-movement frequency — the organ the entire conversation is about is a blind spot. The one GLP-1 colonic-transit trial that exists used liraglutide in 48 people with type 1 diabetes and nerve damage, tracked with a wireless motility capsule, and found large-bowel transit time fell by 31.7 percent. It sped up. That is a narrow population with documented baseline delay and a different drug, so it proves nothing about a healthy person on Wegovy — but it is the only direct evidence anyone has, and it points the opposite way from the explanation in every article on the subject.

The mechanism that is well evidenced is duller and more obvious: you eat less. Semaglutide cut total energy intake by 24 percent across ad libitum meals in a 30-person study. Less going in means less coming out, on a straightforward arithmetic that requires no motility theory at all. Whether fibre specifically falls has never actually been measured — that step is a reasonable inference rather than a finding, though it is the one most worth acting on.

How longThe 47-day number, and the half of it that goes missing

You will see it stated that constipation on semaglutide lasts a median of 47 days. That figure is real and it comes from a pooled analysis of the STEP 1 to 3 trials, covering 2,117 people on semaglutide against 1,262 on placebo. In the same analysis, nausea ran a median of 8 days, diarrhoea 3, vomiting 2.

The part that never travels with it: constipation in the placebo group lasted a median of 35 days. The difference attributable to the drug is therefore about twelve days, not forty-seven. Quoting the raw number without its comparator overstates the drug’s contribution roughly fourfold, and it is quoted without its comparator almost everywhere. Worth adding from the same paper: 98.1 percent of gastrointestinal events were mild to moderate and 99.5 percent were non-serious, and they clustered during dose escalation rather than persisting indefinitely.

The paradoxCommon on these drugs, rarely caused by them

The most interesting recent finding turns the framing inside out. Researchers analysing NIH All of Us data compared 11,656 GLP-1 users with 91,305 non-users, all with obesity, then followed 8,188 matched pairs over time. Constipation was more common among GLP-1 users — 31.6 percent against 26.8 — exactly as the discourse would predict.

But new constipation, appearing after the drug started, ran at just 3.34 cases per 100 person-years: 2.9 on semaglutide, 4.16 on liraglutide, 1.81 on tirzepatide, arriving at a mean of 141 days. As the study’s senior author put it, the most intriguing finding was the contrast between high prevalence and low incidence. Most of the constipation among GLP-1 users was already there. It is a study of health records rather than a trial, and obesity carries its own association with constipation independently — which is precisely the point being made. A drug that a lot of constipated people take is not the same thing as a drug that makes people constipated.

The practical readWhat the evidence supports

Assembling it: there is no GLP-1 frequency to hit, and the population range of three a day to three a week does not change because a prescription did. Both constipation and diarrhoea are on every label, they are more common than on placebo, and they are not rare — but they are mostly mild, mostly clustered around dose increases, and on tirzepatide they trade places as the dose climbs. The best-evidenced reason for things slowing is simply that far less food is going in. And a good share of the constipation on these drugs was there before them.

The lever with the most evidence behind it is the least pharmacological one: when intake drops by a quarter, fibre drops with it unless something is done deliberately, and our fibre calculator is a reasonable way to see where you actually land. Anything persistent, anything that changes sharply, and any decision about a dose belongs with the prescriber who started it — they have seen all of this before, it is on the label they prescribed from, and none of it is a reason to change a treatment on your own. This article is educational, reports what the cited labels and studies measured, and is not medical advice.