No study has tested whether stopping Ozempic improves the bowel prep for a colonoscopy. What has been measured is the prep in people still taking a GLP-1 drug. In one matched cohort, inadequate prep was recorded for 125 of 503 GLP-1 users and 56 of 503 non-users [1]. A repeat procedure followed for 26 against 9. Pooled studies disagree on how large the effect is.
The worry follows from how the drug works. GLP-1 drugs slow the stomach, and a colonoscopy depends on a gut that has emptied. For the upper scope, the concern is food left in the stomach under sedation. That is covered in the gastric ultrasound study of residual stomach contents. This page is about the lower scope, where the concern is a colon the camera cannot see.
What a single large cohort found
A multicenter Veterans Affairs cohort matched 503 GLP-1 users with 503 non-users [1]. The groups were matched for diabetes and cirrhosis. The main outcome combined several signs of inadequate prep, taken from procedure reports.
Users had higher odds of that outcome: 125 against 56, an odds ratio of 2.6 (95% CI 1.9 to 3.7). They also had higher odds of a repeat procedure: 26 against 9, an odds ratio of 3.0. The study was retrospective. It cannot say whether the drug, the diabetes behind it, or something else explains the gap.
Four meta-analyses, no agreement
The pooled evidence does not agree with itself. That disagreement is the finding.
The first pooled 5 studies and 10,833 patients [2]. Inadequate prep was more likely on a GLP-1 drug, at an odds ratio of 2.10 (95% CI 1.41 to 3.13). The mean prep score was lower by 0.34 points.
The second pooled 12 studies and 123,858 patients, including conference abstracts [3]. The unadjusted odds ratio was 1.55 (95% CI 1.11 to 2.16). Adjusted for age, sex, body mass index and diabetes, it was 2.35. It found no association with repeat colonoscopy (OR 1.5, 95% CI 0.88 to 2.56).
The third pooled 6 studies and found no difference [4]. Its odds ratio was 1.00 (95% CI 0.73 to 1.37). Its authors advised against stopping GLP-1 drugs before colonoscopy, and asked for more evidence.
A fourth pooled 8,778 users and 8,290 controls [6]. It put the gap in absolute terms: about 6 percentage points more inadequate prep on a GLP-1 drug. It named the washout period before colonoscopy as the question still to be answered.
Does worse prep mean missed polyps?
That is what a poor prep risks, so one cohort checked it directly [5]. It included 49,987 patients having outpatient screening or surveillance colonoscopy, 4,269 of them GLP-1 users. Users had modestly higher odds of inadequate prep, at an odds ratio of 1.23.
Adenoma detection did not differ. The subgroup with diabetes had the largest prep gap, at an odds ratio of 1.88. It also had lower detection of sessile serrated polyps, at 0.71. The authors called for bowel-prep protocols tailored to this group, tested prospectively.
Observational procedure studies need care in both directions. A cohort can also make a drug look protective across every outcome, as in the ERCP cohort of 21,818.
The one randomized trial nearby
OCULUS randomized people on a GLP-1 or GLP-1/GIP drug before an upper endoscopy [7]. They either held one dose or kept taking it. It was stopped early after 60 patients. Clinically significant residual stomach contents occurred in 3.1% who held against 25.0% who continued.
Its colonoscopy finding is one subgroup. Among 25 people having both scopes, all on clear liquids the day before, none had clinically significant stomach contents. The authors suggest clear liquids may reduce that risk whether or not the drug is taken. The trial did not measure bowel prep quality. Aspiration itself is covered in whether GLP-1 drugs raise aspiration risk.
What this leaves the person booked for a scope
The endoscopy team sets the prep and any pause, and needs to know the drug, dose and date of the last injection. The evidence supports telling them. It does not support a particular number of days off the drug. The bowel-prep studies report GLP-1 drugs as a class, not Ozempic alone. Compounded semaglutide is not FDA-approved and was not studied as such. Slow emptying as a lasting problem is a separate question, set out in whether GLP-1 drugs cause gastroparesis.