How do I distinguish drug-related constipation from something obstructive, given both present as not going?

I have had slow bowels since starting semaglutide and I have accepted that as the trade. Last week was different and it frightened me: four days with nothing, a hard swollen abdomen, no appetite at all, and I vomited twice on the fourth day. It resolved after a very unpleasant night and I am now back to my usual sluggish baseline.

Reading afterwards, I found that ileus and bowel obstruction are listed as reported events on this class, which I had not known. What I now cannot work out is whether what happened to me was an ordinary bad constipation episode or the beginning of something that would have become serious if it had continued.

The reason I am asking rather than assuming: I do not want to go to an emergency department every time I have a slow week, and I equally do not want to sit at home rationalising a genuine obstruction because I have been told constipation is expected on this drug. What are the features that separate them, and what would you not wait on?

The discriminating combination is vomiting plus distension plus absence of both stool and flatus, with pain that is colicky or severe. Any three of those together is an assessment, not a management-at-home situation. Your episode had three of the four, which is why it frightened you and why the instinct was correct even though it resolved.

Being direct about the risk framing first, because it matters more than the taxonomy: mechanical obstruction and ileus have both been reported on this class, ileus appears in product labelling for some of these agents following pharmacovigilance review, and the delayed gastric emptying that produces the ordinary symptoms is the same physiology that makes a functional obstruction plausible. The absolute risk is low. The consequence of missing one is not.

The distinguishing features

What I would not wait on

Any of these is an urgent assessment on the same day, and the middle three warrant an emergency department rather than a routine appointment:

• No flatus and no stool for more than about 24 hours together with vomiting.
• A tense, distended abdomen that is progressively worsening.
• Severe or constant abdominal pain, particularly with tenderness, guarding, fever or tachycardia. Those suggest complications rather than uncomplicated obstruction.
• Vomiting that prevents any fluid intake.
• Vomitus that is bilious, faeculent-smelling, or contains blood.
• Any of the above in someone with previous abdominal surgery, adhesions, a hernia, inflammatory bowel disease or known strictures, all of which raise the baseline mechanical risk substantially.

On your episode

Four days without stool, hard distension and vomiting on day four is a presentation that warranted assessment at the time. The fact that it resolved does not retrospectively make it benign; the majority of ileus episodes resolve conservatively, and "it got better" is the usual outcome rather than evidence that nothing was happening. The feature I would want to know, and which your account does not include, is whether you were passing flatus. If you were, that is substantially reassuring. If you were not, that was an episode of at least partial functional obstruction and it should be on your medical record.

Two things worth doing rather than nothing:

• Report it to whoever prescribes for you, describing the four features rather than the label. "Four days, no stool, distended, vomited twice" gets a different response from "I had bad constipation".
• Treat it as an indication that the current combination of dose and bowel management is not adequate, because an episode like that is more likely to recur than not. The management is preventive: reliable osmotic maintenance, fluid, and a lower threshold for acting next time.

Why this class specifically

The delayed emptying is not confined to the stomach. Small-bowel and colonic transit both slow. Add reduced oral intake, dehydration, opioid use in anyone taking them for anything, and often a period of vomiting, and you have assembled most of the standard risk factors for functional obstruction at once. That is why ileus appears in the pharmacovigilance data for this class rather than being a random association, and it is also why the preventive measures are the same boring ones: fluid, an osmotic agent used reliably rather than reactively, and not ignoring a colon that has stopped entirely.

Nothing here is a substitute for being examined. The purpose of the table is to help you decide whether to be examined, not to replace it.

edited 28 Apr 2026 by mz_4113 — expanded the table to cover the lower concentration

A category the table does not separate out and which accounts for a fair number of episodes like the one described: faecal impaction with overflow. It sits between ordinary constipation and obstruction, it is common in slow transit, and it is frequently misread in both directions.

What happens: a hard mass forms in the rectum or distal colon, too firm to pass. Liquid content from upstream tracks around it. The person then reports diarrhoea, or intermittent leakage, and concludes their constipation has resolved. Meanwhile the impaction is still there, distension continues, and appetite disappears.

Features that suggest it:

• An alternating pattern of nothing for days followed by loose or liquid stool, which reads as irritable bowel but is not.
• A persistent sensation of incomplete evacuation or of needing to go without being able to.
• Leakage or soiling, which people are reluctant to mention and which is diagnostically important.
• Distension and anorexia disproportionate to the reported stool frequency.
• Oral laxatives producing liquid output and cramping without relieving the distension.

Why it matters practically: oral osmotic agents alone are a slow and uncomfortable way to clear a distal impaction, because you are softening content upstream of a plug. The distal component usually needs addressing rectally, and that is worth an actual examination rather than escalating oral doses, which is the common self-managed error and which produces a great deal of cramping and leakage for little benefit.

The prevention is the same as everything else in this topic: do not let it get to the point where a mass has time to dry out. A colon on a four-day cycle with reduced fluid is well set up to form one. Reliable maintenance rather than reactive rescue is the whole strategy, and the reason reactive rescue underperforms is that by the time you are reaching for it, the content has already spent three days losing water.

Worth adding the incidence context, since the question is partly about how much weight to give a rare event.

Constipation as an adverse event is common and well quantified: 23.4% on semaglutide 2.4 mg against 9.5% on placebo in STEP 1 [1], and 11.7-17.1% across the tirzepatide arms against 5.8% on placebo in SURMOUNT-1 [2]. Note the placebo columns again: something like 6-10% of people report constipation over a year on nothing at all, so the drug-attributable increment is closer to 6-14 percentage points than to the headline.

Ileus and obstruction are a completely different order of frequency. They do not appear as quantified rows in the pivotal safety tables, because they were too rare to produce a meaningful cell in trials of a few thousand participants. What exists instead is:

• Post-marketing pharmacovigilance reports, which resulted in ileus being added to labelling for some agents in this class. Spontaneous reporting cannot give you an incidence rate, only a signal.
• Observational pharmacoepidemiology, which has produced mixed results, some studies finding an association with bowel obstruction and others not, with the usual confounding by indication and by prior abdominal surgery.

So the honest statement is: common symptom, rare complication, no reliable incidence figure for the complication, and a plausible mechanistic link rather than a demonstrated causal rate. That combination is exactly the situation in which pattern recognition matters more than probability estimates. You cannot compute your risk. You can recognise the presentation, which is why the table in the accepted answer is more useful than a number would be.

One additional practical note on a group at elevated baseline risk that gets overlooked: anyone with previous abdominal or pelvic surgery has adhesions, and adhesions are the commonest cause of small-bowel obstruction in the general population. Adding a drug that slows transit to an abdomen with adhesions is a different proposition from adding it to one without, and it is worth mentioning in the initial conversation rather than after an episode.