Ozempic and Constipation: The 111,545-Person Answer

2026-09-02 · By the Diet for Constipation data editorial team · 1,628 words

One of the most common questions that arrives alongside a GLP-1 prescription — semaglutide (Ozempic, Wegovy), tirzepatide (Zepbound, Mounjaro), liraglutide (Saxenda) — is not about the injection itself. It is about the bathroom. Nausea gets the headlines, but constipation is the side effect that tends to linger, quietly, in the background of daily life. How likely is it, really? Until recently, the honest answer was: we had trial numbers, but not much real-world data.

That gap just got smaller. A research letter published in Clinical Gastroenterology and Hepatology in 2026, led by Madeleine Chang of Harvard University, analyzed records from the All of Us Research Program — more than 111,000 adults with obesity — to ask two separate questions: how common is constipation among people taking GLP-1 drugs, and how often does the drug actually cause new constipation in people who never had it? The two answers turned out to be quite different, and the difference matters for anyone weighing a start, a continuation, or a dose change.

How the cohort was split in two

The team used the All of Us database, a large and diverse US research cohort, and split the analysis in two. The first part was cross-sectional: compare constipation rates between GLP-1 users and non-users across the whole study population of 111,545 adults with obesity — including 11,656 GLP-1 users, plus smaller groups on older weight-loss medications (bupropion-naltrexone and phentermine-topiramate) for context. The second part was longitudinal and more demanding: take only people with no prior constipation, follow them for up to one year after starting a GLP-1, and see how many develop it compared with matched non-users. For that arm, 8,188 GLP-1 users were matched against non-users — 6,080 on semaglutide, 3,126 on liraglutide, and 940 on tirzepatide.

The two findings, side by side

The cross-sectional half confirmed what many people suspect: constipation is more common among GLP-1 users than non-users — 31.6% versus 26.8%. That is a real gap, and it was statistically solid. But notice what the raw numbers also show: even among people not taking these drugs, more than one in four adults with obesity report constipation. The baseline is already high. The medication adds to a crowded field.

The longitudinal half is where the study earns its keep. Among people who started a GLP-1 with no constipation history, the one-year rate of new-onset constipation was 3.34 per 100 person-years, with the average event arriving around 141 days in. Yes, the relative risk was elevated — a hazard ratio of 1.49 compared with non-users, meaning roughly a 50% higher relative chance. And yes, there were differences between drugs: semaglutide came in at 1.31, liraglutide at 1.44, and tirzepatide at 2.83. But here is the number that should reframe the conversation: fewer than 4% of people developed new constipation within the first year. A 50% relative increase on a small base is still a small number.

The authors put it plainly: new-onset constipation should be monitored, particularly in people with other health conditions, but concerns about it should not be considered a major reason to hold back from starting GLP-1 therapy. For the vast majority of people, this is a manageable side effect, not a dealbreaker.

Why GLP-1 drugs back things up

The mechanism is not mysterious. GLP-1 medications work partly by slowing gastric emptying — food lingers in the stomach longer, which is a feature, not a bug, because the fullness is the point. But the same deceleration can extend downstream. Add two other forces and the plumbing slows further: people on these drugs simply eat much less, which means less bulk moving through the intestine, and many also drink less, because thirst cues get quieter alongside appetite.

That second point deserves emphasis, because it is where food choices become the main lever a person actually controls. The drug shrinks your appetite; it does not shrink your fiber and water needs. Someone eating a third of their previous volume needs each bite to carry more fiber per calorie than before, not less. Our own database makes the stakes visible: white bread delivers 2.7 g of fiber per 100 g while whole-wheat bread delivers 6.0 g; white rice manages 0.4 g while quinoa delivers 7.0 g. When total intake collapses, the gap between a refined-grain diet and a whole-grain one stops being a nutrition footnote and becomes the difference between regular and backed up.

The trial label numbers tell the same story at the population level. In the Wegovy (semaglutide 2.4 mg) trials, constipation was reported by 24% of participants versus 11% on placebo — roughly double the rate, but still a minority. In the Ozempic trials at diabetes doses, the rates were much lower (3-5%). Dose, population and reporting thresholds all move these numbers around, which is exactly why the new real-world analysis is useful: it reflects what happens outside carefully screened trial populations.

What actually helps, according to the practical evidence

The standard playbook for constipation on a GLP-1 is the same playbook we rate foods by across this site — with one adjustment: go slower than you think. Rapidly doubling fiber on a stomach that is already emptying slowly invites bloating and discomfort. The sequence that tends to work:

• Water first. Appetite loss quietly shrinks fluid intake too. A fiber increase without matching water makes stool harder, not softer. This is the single most common unforced error.
• Fiber density over fiber volume. When portions are small, pick foods that earn their space: legumes, oat bran, whole-wheat bread, pears and kiwifruit all clear 3 g of fiber per 100 g — the green threshold our engine uses.
• Distribute it. Three small fiber anchors across the day beat one heroic salad, both for tolerance and for keeping the gut's bacteria supplied with their preferred fuel.
• Keep moving. Walking is one of the few free, well-supported transit-time interventions. On a medication that slows the gut, movement works in the opposite direction.
• Mention it. Persistent constipation on a GLP-1 is a conversation with your care team, not a private endurance contest. Dose timing, formulation and supportive over-the-counter options all exist, and the study's own conclusion was that this side effect is manageable.

Four limits the authors name themselves

Four limitations, stated clearly. First, this was a research letter built on electronic health records from a voluntary cohort — mild constipation handled with over-the-counter products may simply never have been recorded, which would push the incidence estimate in either direction. Second, the study used standardized condition codes rather than validated symptom scores, so severity is unknown. Third, people who already had constipation before starting the drug were excluded from the longitudinal arm — meaning the study cannot say how often GLP-1s worsen existing constipation, only how often they create new cases. Fourth, observational matching reduces but never eliminates confounding: people who choose GLP-1s differ from people who do not in ways a database cannot fully absorb.

None of these caveats sinks the core message. If anything, the direction of every limitation biases toward undercounting — which makes the "under 4% in year one" finding more reassuring, not less.

How we would put this to work

If you are starting a GLP-1, approach the first four months — the window when the average new case appeared in this study — as a period of deliberate fiber and fluid planning rather than an afterthought. Pick two or three green-rated foods you genuinely like, make them daily habits, and keep water visible. If you are already constipated on a GLP-1, the food-first sequence above still applies before anything stronger, and the conversation with your care team is worth having sooner rather than later.

And if you are one of the majority who sails through without digestive trouble: the study explains that too. Small base rates stay small. The bathroom question now has a real-world answer, and it is a calmer one than the internet's would suggest.

Sources

• Chang M, et al. New-onset constipation with GLP-1 receptor agonist use in adults with obesity — research letter, Clinical Gastroenterology and Hepatology, 2026 (All of Us Research Program cohort, N=111,545).
• Medscape, "Constipation Common but Rarely New in GLP-1 Users," 2026.
• FDA label information: Wegovy (semaglutide 2.4 mg) and Ozempic adverse-reaction tables.
• NIDDK: Eating, Diet & Nutrition for Constipation.

FAQs

→ Related: Best Foods for Constipation

This article is informational only, based on published research and USDA data. It is maintained by our data editorial team — not medical professionals. Nothing here is medical advice.